Merge branch 'feature/procedure_v2' into brandon/cre-107-pipeline-conditional-routing

Clean up docs
Clean up end of docs
2025-12-24 16:28:29 +00:00 · 2024-07-31 11:31:30 -04:00 · 2024-07-31 10:58:26 -04:00 · 2024-07-31 10:57:09 -04:00 · 2024-07-31 10:54:20 -04:00 · 2024-07-31 10:12:37 -04:00
21 changed files with 1707 additions and 287 deletions
--- a/docs/core-concepts/Pipeline.md
+++ b/docs/core-concepts/Pipeline.md
@@ -0,0 +1,267 @@
+---
+title: crewAI Pipelines
+description: Understanding and utilizing pipelines in the crewAI framework for efficient multi-stage task processing.
+---
+
+## What is a Pipeline?
+
+A pipeline in crewAI represents a structured workflow that allows for the sequential or parallel execution of multiple crews. It provides a way to organize complex processes involving multiple stages, where the output of one stage can serve as input for subsequent stages.
+
+## Key Terminology
+
+Understanding the following terms is crucial for working effectively with pipelines:
+
+- **Stage**: A distinct part of the pipeline, which can be either sequential (a single crew) or parallel (multiple crews executing concurrently).
+- **Run**: A specific execution of the pipeline for a given set of inputs, representing a single instance of processing through the pipeline.
+- **Branch**: Parallel executions within a stage (e.g., concurrent crew operations).
+- **Trace**: The journey of an individual input through the entire pipeline, capturing the path and transformations it undergoes.
+
+Example pipeline structure:
+
+```
+crew1 >> [crew2, crew3] >> crew4
+```
+
+This represents a pipeline with three stages:
+
+1. A sequential stage (crew1)
+2. A parallel stage with two branches (crew2 and crew3 executing concurrently)
+3. Another sequential stage (crew4)
+
+Each input creates its own run, flowing through all stages of the pipeline. Multiple runs can be processed concurrently, each following the defined pipeline structure.
+
+## Pipeline Attributes
+
+| Attribute  | Parameters | Description                                                                           |
+| :--------- | :--------- | :------------------------------------------------------------------------------------ |
+| **Stages** | `stages`   | A list of crews or lists of crews representing the stages to be executed in sequence. |
+
+## Creating a Pipeline
+
+When creating a pipeline, you define a series of stages, each consisting of either a single crew or a list of crews for parallel execution. The pipeline ensures that each stage is executed in order, with the output of one stage feeding into the next.
+
+### Example: Assembling a Pipeline
+
+```python
+from crewai import Crew, Agent, Task, Pipeline
+
+# Define your crews
+research_crew = Crew(
+    agents=[researcher],
+    tasks=[research_task],
+    process=Process.sequential
+)
+
+analysis_crew = Crew(
+    agents=[analyst],
+    tasks=[analysis_task],
+    process=Process.sequential
+)
+
+writing_crew = Crew(
+    agents=[writer],
+    tasks=[writing_task],
+    process=Process.sequential
+)
+
+# Assemble the pipeline
+my_pipeline = Pipeline(
+    stages=[research_crew, analysis_crew, writing_crew]
+)
+```
+
+## Pipeline Methods
+
+| Method           | Description                                                                                                                                                                    |
+| :--------------- | :----------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| **process_runs** | Executes the pipeline, processing all stages and returning the results. This method initiates one or more runs through the pipeline, handling the flow of data between stages. |
+
+## Pipeline Output
+
+!!! note "Understanding Pipeline Outputs"
+The output of a pipeline in the crewAI framework is encapsulated within two main classes: `PipelineOutput` and `PipelineRunResult`. These classes provide a structured way to access the results of the pipeline's execution, including various formats such as raw strings, JSON, and Pydantic models.
+
+### Pipeline Output Attributes
+
+| Attribute       | Parameters    | Type                      | Description                                                                                               |
+| :-------------- | :------------ | :------------------------ | :-------------------------------------------------------------------------------------------------------- |
+| **ID**          | `id`          | `UUID4`                   | A unique identifier for the pipeline output.                                                              |
+| **Run Results** | `run_results` | `List[PipelineRunResult]` | A list of `PipelineRunResult` objects, each representing the output of a single run through the pipeline. |
+
+### Pipeline Output Methods
+
+| Method/Property    | Description                                            |
+| :----------------- | :----------------------------------------------------- |
+| **add_run_result** | Adds a `PipelineRunResult` to the list of run results. |
+
+### Pipeline Run Result Attributes
+
+| Attribute         | Parameters      | Type                       | Description                                                                                   |
+| :---------------- | :-------------- | :------------------------- | :-------------------------------------------------------------------------------------------- |
+| **ID**            | `id`            | `UUID4`                    | A unique identifier for the run result.                                                       |
+| **Raw**           | `raw`           | `str`                      | The raw output of the final stage in the pipeline run.                                        |
+| **Pydantic**      | `pydantic`      | `Optional[BaseModel]`      | A Pydantic model object representing the structured output of the final stage, if applicable. |
+| **JSON Dict**     | `json_dict`     | `Optional[Dict[str, Any]]` | A dictionary representing the JSON output of the final stage, if applicable.                  |
+| **Token Usage**   | `token_usage`   | `Dict[str, Any]`           | A summary of token usage across all stages of the pipeline run.                               |
+| **Trace**         | `trace`         | `List[Any]`                | A trace of the journey of inputs through the pipeline run.                                    |
+| **Crews Outputs** | `crews_outputs` | `List[CrewOutput]`         | A list of `CrewOutput` objects, representing the outputs from each crew in the pipeline run.  |
+
+### Pipeline Run Result Methods and Properties
+
+| Method/Property | Description                                                                                              |
+| :-------------- | :------------------------------------------------------------------------------------------------------- |
+| **json**        | Returns the JSON string representation of the run result if the output format of the final task is JSON. |
+| **to_dict**     | Converts the JSON and Pydantic outputs to a dictionary.                                                  |
+| \***\*str\*\*** | Returns the string representation of the run result, prioritizing Pydantic, then JSON, then raw.         |
+
+### Accessing Pipeline Outputs
+
+Once a pipeline has been executed, its output can be accessed through the `PipelineOutput` object returned by the `process_runs` method. The `PipelineOutput` class provides access to individual `PipelineRunResult` objects, each representing a single run through the pipeline.
+
+#### Example
+
+```python
+# Define input data for the pipeline
+input_data = [{"initial_query": "Latest advancements in AI"}, {"initial_query": "Future of robotics"}]
+
+# Execute the pipeline
+pipeline_output = await my_pipeline.process_runs(input_data)
+
+# Access the results
+for run_result in pipeline_output.run_results:
+    print(f"Run ID: {run_result.id}")
+    print(f"Final Raw Output: {run_result.raw}")
+    if run_result.json_dict:
+        print(f"JSON Output: {json.dumps(run_result.json_dict, indent=2)}")
+    if run_result.pydantic:
+        print(f"Pydantic Output: {run_result.pydantic}")
+    print(f"Token Usage: {run_result.token_usage}")
+    print(f"Trace: {run_result.trace}")
+    print("Crew Outputs:")
+    for crew_output in run_result.crews_outputs:
+        print(f"  Crew: {crew_output.raw}")
+    print("\n")
+```
+
+This example demonstrates how to access and work with the pipeline output, including individual run results and their associated data.
+
+## Using Pipelines
+
+Pipelines are particularly useful for complex workflows that involve multiple stages of processing, analysis, or content generation. They allow you to:
+
+1. **Sequence Operations**: Execute crews in a specific order, ensuring that the output of one crew is available as input to the next.
+2. **Parallel Processing**: Run multiple crews concurrently within a stage for increased efficiency.
+3. **Manage Complex Workflows**: Break down large tasks into smaller, manageable steps executed by specialized crews.
+
+### Example: Running a Pipeline
+
+```python
+# Define input data for the pipeline
+input_data = [{"initial_query": "Latest advancements in AI"}]
+
+# Execute the pipeline, initiating a run for each input
+results = await my_pipeline.process_runs(input_data)
+
+# Access the results
+for result in results:
+    print(f"Final Output: {result.raw}")
+    print(f"Token Usage: {result.token_usage}")
+    print(f"Trace: {result.trace}")  # Shows the path of the input through all stages
+```
+
+## Advanced Features
+
+### Parallel Execution within Stages
+
+You can define parallel execution within a stage by providing a list of crews, creating multiple branches:
+
+```python
+parallel_analysis_crew = Crew(agents=[financial_analyst], tasks=[financial_analysis_task])
+market_analysis_crew = Crew(agents=[market_analyst], tasks=[market_analysis_task])
+
+my_pipeline = Pipeline(
+    stages=[
+        research_crew,
+        [parallel_analysis_crew, market_analysis_crew],  # Parallel execution (branching)
+        writing_crew
+    ]
+)
+```
+
+### Routers in Pipelines
+
+Routers are a powerful feature in crewAI pipelines that allow for dynamic decision-making and branching within your workflow. They enable you to direct the flow of execution based on specific conditions or criteria, making your pipelines more flexible and adaptive.
+
+#### What is a Router?
+
+A router in crewAI is a special component that can be included as a stage in your pipeline. It evaluates the input data and determines which path the execution should take next. This allows for conditional branching in your pipeline, where different crews or sub-pipelines can be executed based on the router's decision.
+
+#### Key Components of a Router
+
+1. **Routes**: A dictionary of named routes, each associated with a condition and a pipeline to execute if the condition is met.
+2. **Default Route**: A fallback pipeline that is executed if none of the defined route conditions are met.
+
+#### Creating a Router
+
+Here's an example of how to create a router:
+
+```python
+from crewai import Router, Route, Pipeline, Crew, Agent, Task
+
+# Define your agents
+classifier = Agent(name="Classifier", role="Email Classifier")
+urgent_handler = Agent(name="Urgent Handler", role="Urgent Email Processor")
+normal_handler = Agent(name="Normal Handler", role="Normal Email Processor")
+
+# Define your tasks
+classify_task = Task(description="Classify the email based on its content and metadata.")
+urgent_task = Task(description="Process and respond to urgent email quickly.")
+normal_task = Task(description="Process and respond to normal email thoroughly.")
+
+# Define your crews
+classification_crew = Crew(agents=[classifier], tasks=[classify_task]) # classify email between high and low urgency 1-10
+urgent_crew = Crew(agents=[urgent_handler], tasks=[urgent_task])
+normal_crew = Crew(agents=[normal_handler], tasks=[normal_task])
+
+# Create pipelines for different urgency levels
+urgent_pipeline = Pipeline(stages=[classification_crew, urgent_crew])
+normal_pipeline = Pipeline(stages=[classification_crew, normal_crew])
+
+# Create a router
+email_router = Router(
+    routes={
+        "high_urgency": Route(
+            condition=lambda x: x.get("urgency_score", 0) > 7,
+            pipeline=urgent_pipeline
+        ),
+        "low_urgency": Route(
+            condition=lambda x: x.get("urgency_score", 0) <= 7,
+            pipeline=normal_pipeline
+        )
+    },
+    default=Pipeline(stages=[normal_pipeline])  # Default to just classification if no urgency score
+)
+
+# Use the router in a main pipeline
+main_pipeline = Pipeline(stages=[email_router])
+
+inputs = [{"email": "..."}, {"email": "..."}]  # List of email data
+
+main_pipeline.kickoff(inputs=inputs)
+```
+
+In this example, the router decides between an urgent pipeline and a normal pipeline based on the urgency score of the email. If the urgency score is greater than 7, it routes to the urgent pipeline; otherwise, it uses the normal pipeline. If the input doesn't include an urgency score, it defaults to just the classification crew.
+
+#### Benefits of Using Routers
+
+1. **Dynamic Workflow**: Adapt your pipeline's behavior based on input characteristics or intermediate results.
+2. **Efficiency**: Route urgent tasks to quicker processes, reserving more thorough pipelines for less time-sensitive inputs.
+3. **Flexibility**: Easily modify or extend your pipeline's logic without changing the core structure.
+4. **Scalability**: Handle a wide range of email types and urgency levels with a single pipeline structure.
+
+### Error Handling and Validation
+
+The Pipeline class includes validation mechanisms to ensure the robustness of the pipeline structure:
+
+- Validates that stages contain only Crew instances or lists of Crew instances.
+- Prevents double nesting of stages to maintain a clear structure.
--- a/docs/index.md
+++ b/docs/index.md
@@ -46,6 +46,11 @@ Cutting-edge framework for orchestrating role-playing, autonomous AI agents. By
                    Crews
                </a>
            </li>
+            <li>
+                <a href="./core-concepts/Pipeline">
+                    Pipeline
+                </a>
+            </li>
            <li>
                <a href="./core-concepts/Training-Crew">
                    Training
--- a/poetry.lock
+++ b/poetry.lock
@@ -1,14 +1,14 @@
-# This file is automatically @generated by Poetry 1.8.3 and should not be changed by hand.
+# This file is automatically @generated by Poetry 1.7.1 and should not be changed by hand.

 [[package]]
 name = "agentops"
-version = "0.3.2"
+version = "0.3.4"
 description = "Python SDK for developing AI agent evals and observability"
 optional = true
 python-versions = ">=3.7"
 files = [
-    {file = "agentops-0.3.2-py3-none-any.whl", hash = "sha256:b35988e04378624204572bb3d7a454094f879ea573f05b57d4e75ab0bfbb82af"},
-    {file = "agentops-0.3.2.tar.gz", hash = "sha256:55559ac4a43634831dfa8937c2597c28e332809dc7c6bb3bc3c8b233442e224c"},
+    {file = "agentops-0.3.4-py3-none-any.whl", hash = "sha256:126f7aed4ba43c1399b5488d67a03d10cb4c531e619c650776f826ca00c1aa24"},
+    {file = "agentops-0.3.4.tar.gz", hash = "sha256:a92c9cb7c511197f0ecb8cb5aca15d35022c15a3d2fd2aaaa34cd7e5dc59393f"},
 ]

 [package.dependencies]
@@ -355,17 +355,17 @@ lxml = ["lxml"]

 [[package]]
 name = "boto3"
-version = "1.34.146"
+version = "1.34.149"
 description = "The AWS SDK for Python"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "boto3-1.34.146-py3-none-any.whl", hash = "sha256:7ec568fb19bce82a70be51f08fddac1ef927ca3fb0896cbb34303a012ba228d8"},
-    {file = "boto3-1.34.146.tar.gz", hash = "sha256:5686fe2a6d1aa1de8a88e9589cdcc33361640d3d7a13da718a30717248886124"},
+    {file = "boto3-1.34.149-py3-none-any.whl", hash = "sha256:11edeeacdd517bda3b7615b754d8440820cdc9ddd66794cc995a9693ddeaa3be"},
+    {file = "boto3-1.34.149.tar.gz", hash = "sha256:f4e6489ba9dc7fb37d53e0e82dbc97f2cb0a4969ef3970e2c88b8f94023ae81a"},
 ]

 [package.dependencies]
-botocore = ">=1.34.146,<1.35.0"
+botocore = ">=1.34.149,<1.35.0"
 jmespath = ">=0.7.1,<2.0.0"
 s3transfer = ">=0.10.0,<0.11.0"

@@ -374,13 +374,13 @@ crt = ["botocore[crt] (>=1.21.0,<2.0a0)"]

 [[package]]
 name = "botocore"
-version = "1.34.146"
+version = "1.34.149"
 description = "Low-level, data-driven core of boto 3."
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "botocore-1.34.146-py3-none-any.whl", hash = "sha256:3fd4782362bd29c192704ebf859c5c8c5189ad05719e391eefe23088434427ae"},
-    {file = "botocore-1.34.146.tar.gz", hash = "sha256:849cb8e54e042443aeabcd7822b5f2b76cb5cfe33fe3a71f91c7c069748a869c"},
+    {file = "botocore-1.34.149-py3-none-any.whl", hash = "sha256:ae6c4be52eeee96f68c116b27d252bab069cd046d61a17cfe8e9da411cf22906"},
+    {file = "botocore-1.34.149.tar.gz", hash = "sha256:2e1eb5ef40102a3d796bb3dd05f2ac5e8fb43fe1ff114b4f6d33153437f5a372"},
 ]

 [package.dependencies]
@@ -1012,13 +1012,13 @@ idna = ">=2.0.0"

 [[package]]
 name = "embedchain"
-version = "0.1.118"
+version = "0.1.119"
 description = "Simplest open source retrieval (RAG) framework"
 optional = false
 python-versions = "<=3.13,>=3.9"
 files = [
-    {file = "embedchain-0.1.118-py3-none-any.whl", hash = "sha256:38ead471df9d9234bf42e6f7a32cab26431d50d6f2f894f18a6cabc0b02bf31a"},
-    {file = "embedchain-0.1.118.tar.gz", hash = "sha256:1fa1e799882a1dc4e63af344595b043f1c1f30fbd59461b6660b1934b85a1e4b"},
+    {file = "embedchain-0.1.119-py3-none-any.whl", hash = "sha256:8ec3e7f139939fa1dc8fda898f8d8d9d31a5abfe08e184b607e38733d863d606"},
+    {file = "embedchain-0.1.119.tar.gz", hash = "sha256:0f4f45e092b7f3192ea6fe82575726532573b1231d7af6c22edc695b701b4223"},
 ]

 [package.dependencies]
@@ -1032,7 +1032,7 @@ langchain = ">0.2,<=0.3"
 langchain-cohere = ">=0.1.4,<0.2.0"
 langchain-community = ">=0.2.6,<0.3.0"
 langchain-openai = ">=0.1.7,<0.2.0"
-mem0ai = ">=0.0.5,<0.0.6"
+mem0ai = ">=0.0.9,<0.0.10"
 openai = ">=1.1.1"
 posthog = ">=3.0.2,<4.0.0"
 pypdf = ">=4.0.1,<5.0.0"
@@ -1061,20 +1061,6 @@ together = ["together (>=1.2.1,<2.0.0)"]
 vertexai = ["langchain-google-vertexai (>=1.0.6,<2.0.0)"]
 weaviate = ["weaviate-client (>=3.24.1,<4.0.0)"]

-[[package]]
-name = "eval-type-backport"
-version = "0.2.0"
-description = "Like `typing._eval_type`, but lets older Python versions use newer typing features."
-optional = false
-python-versions = ">=3.8"
-files = [
-    {file = "eval_type_backport-0.2.0-py3-none-any.whl", hash = "sha256:ac2f73d30d40c5a30a80b8739a789d6bb5e49fdffa66d7912667e2015d9c9933"},
-    {file = "eval_type_backport-0.2.0.tar.gz", hash = "sha256:68796cfbc7371ebf923f03bdf7bef415f3ec098aeced24e054b253a0e78f7b37"},
-]
-
-[package.extras]
-tests = ["pytest"]
-
 [[package]]
 name = "exceptiongroup"
 version = "1.2.2"
@@ -1402,13 +1388,13 @@ requests = ["requests (>=2.20.0,<3.0.0.dev0)"]

 [[package]]
 name = "google-cloud-aiplatform"
-version = "1.59.0"
+version = "1.60.0"
 description = "Vertex AI API client library"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "google-cloud-aiplatform-1.59.0.tar.gz", hash = "sha256:2bebb59c0ba3e3b4b568305418ca1b021977988adbee8691a5bed09b037e7e63"},
-    {file = "google_cloud_aiplatform-1.59.0-py2.py3-none-any.whl", hash = "sha256:549e6eb1844b0f853043309138ebe2db00de4bbd8197b3bde26804ac163ef52a"},
+    {file = "google-cloud-aiplatform-1.60.0.tar.gz", hash = "sha256:782c7f1ec0e77a7c7daabef3b65bfd506ed2b4b1dc2186753c43cd6faf8dd04e"},
+    {file = "google_cloud_aiplatform-1.60.0-py2.py3-none-any.whl", hash = "sha256:5f14159c9575f4b46335027e3ceb8fa57bd5eaa76a07f858105b8c6c034ec0d6"},
 ]

 [package.dependencies]
@@ -1430,8 +1416,8 @@ cloud-profiler = ["tensorboard-plugin-profile (>=2.4.0,<3.0.0dev)", "tensorflow
 datasets = ["pyarrow (>=10.0.1)", "pyarrow (>=14.0.0)", "pyarrow (>=3.0.0,<8.0dev)"]
 endpoint = ["requests (>=2.28.1)"]
 full = ["cloudpickle (<3.0)", "docker (>=5.0.3)", "explainable-ai-sdk (>=1.0.0)", "fastapi (>=0.71.0,<=0.109.1)", "google-cloud-bigquery", "google-cloud-bigquery-storage", "google-cloud-logging (<4.0)", "google-vizier (>=0.1.6)", "httpx (>=0.23.0,<0.25.0)", "immutabledict", "lit-nlp (==0.4.0)", "mlflow (>=1.27.0,<=2.1.1)", "numpy (>=1.15.0)", "pandas (>=1.0.0)", "pandas (>=1.0.0,<2.2.0)", "pyarrow (>=10.0.1)", "pyarrow (>=14.0.0)", "pyarrow (>=3.0.0,<8.0dev)", "pyarrow (>=6.0.1)", "pydantic (<2)", "pyyaml (>=5.3.1,<7)", "ray[default] (>=2.4,<2.5.dev0 || >2.9.0,!=2.9.1,!=2.9.2,<=2.9.3)", "ray[default] (>=2.5,<=2.9.3)", "requests (>=2.28.1)", "setuptools (<70.0.0)", "starlette (>=0.17.1)", "tensorboard-plugin-profile (>=2.4.0,<3.0.0dev)", "tensorflow (>=2.3.0,<3.0.0dev)", "tensorflow (>=2.3.0,<3.0.0dev)", "tensorflow (>=2.4.0,<3.0.0dev)", "tqdm (>=4.23.0)", "urllib3 (>=1.21.1,<1.27)", "uvicorn[standard] (>=0.16.0)", "werkzeug (>=2.0.0,<2.1.0dev)"]
-langchain = ["langchain (>=0.1.16,<0.3)", "langchain-core (<0.2)", "langchain-google-vertexai (<2)", "openinference-instrumentation-langchain (>=0.1.19,<0.2)", "tenacity (<=8.3)"]
-langchain-testing = ["absl-py", "cloudpickle (>=3.0,<4.0)", "langchain (>=0.1.16,<0.3)", "langchain-core (<0.2)", "langchain-google-vertexai (<2)", "openinference-instrumentation-langchain (>=0.1.19,<0.2)", "opentelemetry-exporter-gcp-trace (<2)", "opentelemetry-sdk (<2)", "pydantic (>=2.6.3,<3)", "pytest-xdist", "tenacity (<=8.3)"]
+langchain = ["langchain (>=0.1.16,<0.3)", "langchain-core (<0.3)", "langchain-google-vertexai (<2)", "openinference-instrumentation-langchain (>=0.1.19,<0.2)", "tenacity (<=8.3)"]
+langchain-testing = ["absl-py", "cloudpickle (>=3.0,<4.0)", "google-cloud-trace (<2)", "langchain (>=0.1.16,<0.3)", "langchain-core (<0.3)", "langchain-google-vertexai (<2)", "openinference-instrumentation-langchain (>=0.1.19,<0.2)", "opentelemetry-exporter-gcp-trace (<2)", "opentelemetry-sdk (<2)", "pydantic (>=2.6.3,<3)", "pytest-xdist", "tenacity (<=8.3)"]
 lit = ["explainable-ai-sdk (>=1.0.0)", "lit-nlp (==0.4.0)", "pandas (>=1.0.0)", "tensorflow (>=2.3.0,<3.0.0dev)"]
 metadata = ["numpy (>=1.15.0)", "pandas (>=1.0.0)"]
 pipelines = ["pyyaml (>=5.3.1,<7)"]
@@ -1441,7 +1427,7 @@ private-endpoints = ["requests (>=2.28.1)", "urllib3 (>=1.21.1,<1.27)"]
 rapid-evaluation = ["pandas (>=1.0.0,<2.2.0)", "tqdm (>=4.23.0)"]
 ray = ["google-cloud-bigquery", "google-cloud-bigquery-storage", "immutabledict", "pandas (>=1.0.0,<2.2.0)", "pyarrow (>=6.0.1)", "pydantic (<2)", "ray[default] (>=2.4,<2.5.dev0 || >2.9.0,!=2.9.1,!=2.9.2,<=2.9.3)", "ray[default] (>=2.5,<=2.9.3)", "setuptools (<70.0.0)"]
 ray-testing = ["google-cloud-bigquery", "google-cloud-bigquery-storage", "immutabledict", "pandas (>=1.0.0,<2.2.0)", "pyarrow (>=6.0.1)", "pydantic (<2)", "pytest-xdist", "ray[default] (>=2.4,<2.5.dev0 || >2.9.0,!=2.9.1,!=2.9.2,<=2.9.3)", "ray[default] (>=2.5,<=2.9.3)", "ray[train] (==2.9.3)", "scikit-learn", "setuptools (<70.0.0)", "tensorflow", "torch (>=2.0.0,<2.1.0)", "xgboost", "xgboost-ray"]
-reasoningengine = ["cloudpickle (>=3.0,<4.0)", "opentelemetry-exporter-gcp-trace (<2)", "opentelemetry-sdk (<2)", "pydantic (>=2.6.3,<3)"]
+reasoningengine = ["cloudpickle (>=3.0,<4.0)", "google-cloud-trace (<2)", "opentelemetry-exporter-gcp-trace (<2)", "opentelemetry-sdk (<2)", "pydantic (>=2.6.3,<3)"]
 tensorboard = ["tensorboard-plugin-profile (>=2.4.0,<3.0.0dev)", "tensorflow (>=2.3.0,<3.0.0dev)", "tensorflow (>=2.4.0,<3.0.0dev)", "werkzeug (>=2.0.0,<2.1.0dev)"]
 testing = ["bigframes", "cloudpickle (<3.0)", "docker (>=5.0.3)", "explainable-ai-sdk (>=1.0.0)", "fastapi (>=0.71.0,<=0.109.1)", "google-api-core (>=2.11,<3.0.0)", "google-cloud-bigquery", "google-cloud-bigquery-storage", "google-cloud-logging (<4.0)", "google-vizier (>=0.1.6)", "grpcio-testing", "httpx (>=0.23.0,<0.25.0)", "immutabledict", "ipython", "kfp (>=2.6.0,<3.0.0)", "lit-nlp (==0.4.0)", "mlflow (>=1.27.0,<=2.1.1)", "nltk", "numpy (>=1.15.0)", "pandas (>=1.0.0)", "pandas (>=1.0.0,<2.2.0)", "pyarrow (>=10.0.1)", "pyarrow (>=14.0.0)", "pyarrow (>=3.0.0,<8.0dev)", "pyarrow (>=6.0.1)", "pydantic (<2)", "pyfakefs", "pytest-asyncio", "pytest-xdist", "pyyaml (>=5.3.1,<7)", "ray[default] (>=2.4,<2.5.dev0 || >2.9.0,!=2.9.1,!=2.9.2,<=2.9.3)", "ray[default] (>=2.5,<=2.9.3)", "requests (>=2.28.1)", "requests-toolbelt (<1.0.0)", "scikit-learn", "sentencepiece (>=0.2.0)", "setuptools (<70.0.0)", "starlette (>=0.17.1)", "tensorboard-plugin-profile (>=2.4.0,<3.0.0dev)", "tensorflow (==2.13.0)", "tensorflow (==2.16.1)", "tensorflow (>=2.3.0,<3.0.0dev)", "tensorflow (>=2.3.0,<3.0.0dev)", "tensorflow (>=2.4.0,<3.0.0dev)", "torch (>=2.0.0,<2.1.0)", "torch (>=2.2.0)", "tqdm (>=4.23.0)", "urllib3 (>=1.21.1,<1.27)", "uvicorn[standard] (>=0.16.0)", "werkzeug (>=2.0.0,<2.1.0dev)", "xgboost"]
 tokenization = ["sentencepiece (>=0.2.0)"]
@@ -1756,25 +1742,6 @@ files = [
 backports-strenum = {version = ">=1.3", markers = "python_version < \"3.11\""}
 colorama = ">=0.4"

-[[package]]
-name = "groq"
-version = "0.9.0"
-description = "The official Python library for the groq API"
-optional = false
-python-versions = ">=3.7"
-files = [
-    {file = "groq-0.9.0-py3-none-any.whl", hash = "sha256:d0e46f4ad645504672bb09c8100af3ced3a7db0d5119dc13e4aca535fc455874"},
-    {file = "groq-0.9.0.tar.gz", hash = "sha256:130ed5e35d3acfaab46b9e7a078eeaebf91052f4a9d71f86f87fb319b5fec332"},
-]
-
-[package.dependencies]
-anyio = ">=3.5.0,<5"
-distro = ">=1.7.0,<2"
-httpx = ">=0.23.0,<1"
-pydantic = ">=1.9.0,<3"
-sniffio = "*"
-typing-extensions = ">=4.7,<5"
-
 [[package]]
 name = "grpc-google-iam-v1"
 version = "0.13.1"
@@ -2077,13 +2044,13 @@ files = [

 [[package]]
 name = "huggingface-hub"
-version = "0.24.0"
+version = "0.24.3"
 description = "Client library to download and publish models, datasets and other repos on the huggingface.co hub"
 optional = false
 python-versions = ">=3.8.0"
 files = [
-    {file = "huggingface_hub-0.24.0-py3-none-any.whl", hash = "sha256:7ad92edefb93d8145c061f6df8d99df2ff85f8379ba5fac8a95aca0642afa5d7"},
-    {file = "huggingface_hub-0.24.0.tar.gz", hash = "sha256:6c7092736b577d89d57b3cdfea026f1b0dc2234ae783fa0d59caf1bf7d52dfa7"},
+    {file = "huggingface_hub-0.24.3-py3-none-any.whl", hash = "sha256:69ecce486dd6cdad69937ba76779e893c224a670a9d947636c1d5cbd049e44d8"},
+    {file = "huggingface_hub-0.24.3.tar.gz", hash = "sha256:bfdc05cc9b64a0e24e8614a44222698799183268f6b68be209aa2df70cff2cde"},
 ]

 [package.dependencies]
@@ -2161,22 +2128,22 @@ files = [

 [[package]]
 name = "importlib-metadata"
-version = "7.1.0"
+version = "8.0.0"
 description = "Read metadata from Python packages"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "importlib_metadata-7.1.0-py3-none-any.whl", hash = "sha256:30962b96c0c223483ed6cc7280e7f0199feb01a0e40cfae4d4450fc6fab1f570"},
-    {file = "importlib_metadata-7.1.0.tar.gz", hash = "sha256:b78938b926ee8d5f020fc4772d487045805a55ddbad2ecf21c6d60938dc7fcd2"},
+    {file = "importlib_metadata-8.0.0-py3-none-any.whl", hash = "sha256:15584cf2b1bf449d98ff8a6ff1abef57bf20f3ac6454f431736cd3e660921b2f"},
+    {file = "importlib_metadata-8.0.0.tar.gz", hash = "sha256:188bd24e4c346d3f0a933f275c2fec67050326a856b9a359881d7c2a697e8812"},
 ]

 [package.dependencies]
 zipp = ">=0.5"

 [package.extras]
-docs = ["furo", "jaraco.packaging (>=9.3)", "jaraco.tidelift (>=1.4)", "rst.linker (>=1.9)", "sphinx (>=3.5)", "sphinx-lint"]
+doc = ["furo", "jaraco.packaging (>=9.3)", "jaraco.tidelift (>=1.4)", "rst.linker (>=1.9)", "sphinx (>=3.5)", "sphinx-lint"]
 perf = ["ipython"]
-testing = ["flufl.flake8", "importlib-resources (>=1.3)", "jaraco.test (>=5.4)", "packaging", "pyfakefs", "pytest (>=6)", "pytest-checkdocs (>=2.4)", "pytest-cov", "pytest-enabler (>=2.2)", "pytest-mypy", "pytest-perf (>=0.9.2)", "pytest-ruff (>=0.2.1)"]
+test = ["flufl.flake8", "importlib-resources (>=1.3)", "jaraco.test (>=5.4)", "packaging", "pyfakefs", "pytest (>=6,!=8.1.*)", "pytest-checkdocs (>=2.4)", "pytest-cov", "pytest-enabler (>=2.2)", "pytest-mypy", "pytest-perf (>=0.9.2)", "pytest-ruff (>=0.2.1)"]

 [[package]]
 name = "importlib-resources"
@@ -2456,19 +2423,19 @@ tests = ["aiohttp", "duckdb", "pandas (>=1.4)", "polars (>=0.19)", "pytest", "py

 [[package]]
 name = "langchain"
-version = "0.2.10"
+version = "0.2.11"
 description = "Building applications with LLMs through composability"
 optional = false
 python-versions = "<4.0,>=3.8.1"
 files = [
-    {file = "langchain-0.2.10-py3-none-any.whl", hash = "sha256:b4fb58c7faf4f4999cfe3325474979a7121a1737dd101655a723a1d957ef0617"},
-    {file = "langchain-0.2.10.tar.gz", hash = "sha256:1f861c1b59ac9c91b02bb0fa58d3adad1c1d0686636872b5b357bbce3ce41d06"},
+    {file = "langchain-0.2.11-py3-none-any.whl", hash = "sha256:5a7a8b4918f3d3bebce9b4f23b92d050699e6f7fb97591e8941177cf07a260a2"},
+    {file = "langchain-0.2.11.tar.gz", hash = "sha256:d7a9e4165f02dca0bd78addbc2319d5b9286b5d37c51d784124102b57e9fd297"},
 ]

 [package.dependencies]
 aiohttp = ">=3.8.3,<4.0.0"
 async-timeout = {version = ">=4.0.0,<5.0.0", markers = "python_version < \"3.11\""}
-langchain-core = ">=0.2.22,<0.3.0"
+langchain-core = ">=0.2.23,<0.3.0"
 langchain-text-splitters = ">=0.2.0,<0.3.0"
 langsmith = ">=0.1.17,<0.2.0"
 numpy = [
@@ -2504,20 +2471,20 @@ langchain-community = ["langchain-community (>=0.2.4)"]

 [[package]]
 name = "langchain-community"
-version = "0.2.9"
+version = "0.2.10"
 description = "Community contributed LangChain integrations."
 optional = false
 python-versions = "<4.0,>=3.8.1"
 files = [
-    {file = "langchain_community-0.2.9-py3-none-any.whl", hash = "sha256:b51d3adf9346a1161c1098917585b9e303cf24e2f5c71f5d232a0504edada5f2"},
-    {file = "langchain_community-0.2.9.tar.gz", hash = "sha256:1e7c180232916cbe35fe00509680dd1f805e32d7c87b5e80b3a9ec8754ecae37"},
+    {file = "langchain_community-0.2.10-py3-none-any.whl", hash = "sha256:9f4d1b5ab7f0b0a704f538e26e50fce45a461da6d2bf6b7b636d24f22fbc088a"},
+    {file = "langchain_community-0.2.10.tar.gz", hash = "sha256:3a0404bad4bd07d6f86affdb62fb3d080a456c66191754d586a409d9d6024d62"},
 ]

 [package.dependencies]
 aiohttp = ">=3.8.3,<4.0.0"
 dataclasses-json = ">=0.5.7,<0.7"
 langchain = ">=0.2.9,<0.3.0"
-langchain-core = ">=0.2.22,<0.3.0"
+langchain-core = ">=0.2.23,<0.3.0"
 langsmith = ">=0.1.0,<0.2.0"
 numpy = [
    {version = ">=1,<2", markers = "python_version < \"3.12\""},
@@ -2530,13 +2497,13 @@ tenacity = ">=8.1.0,<8.4.0 || >8.4.0,<9.0.0"

 [[package]]
 name = "langchain-core"
-version = "0.2.22"
+version = "0.2.24"
 description = "Building applications with LLMs through composability"
 optional = false
 python-versions = "<4.0,>=3.8.1"
 files = [
-    {file = "langchain_core-0.2.22-py3-none-any.whl", hash = "sha256:7731a86440c0958b3186c003fb9b26b2d5a682a6344bda7bfb9174e2898f8b43"},
-    {file = "langchain_core-0.2.22.tar.gz", hash = "sha256:582d6f929a43b830139444e4124123cd415331ad62f25757b1406252958cdcac"},
+    {file = "langchain_core-0.2.24-py3-none-any.whl", hash = "sha256:9444fc082d21ef075d925590a684a73fe1f9688a3d90087580ec929751be55e7"},
+    {file = "langchain_core-0.2.24.tar.gz", hash = "sha256:f2e3fa200b124e8c45d270da9bf836bed9c09532612c96ff3225e59b9a232f5a"},
 ]

 [package.dependencies]
@@ -2552,13 +2519,13 @@ tenacity = ">=8.1.0,<8.4.0 || >8.4.0,<9.0.0"

 [[package]]
 name = "langchain-experimental"
-version = "0.0.62"
+version = "0.0.63"
 description = "Building applications with LLMs through composability"
 optional = false
 python-versions = "<4.0,>=3.8.1"
 files = [
-    {file = "langchain_experimental-0.0.62-py3-none-any.whl", hash = "sha256:9240f9e3490e819976f20a37863970036e7baacb7104b9eb6833d19ab6d518c9"},
-    {file = "langchain_experimental-0.0.62.tar.gz", hash = "sha256:9737fbc8429d24457ea4d368e3c9ba9ed1cace0564fb5f1a96a3027a588bd0ac"},
+    {file = "langchain_experimental-0.0.63-py3-none-any.whl", hash = "sha256:cb4ae7a685bb3c077d138b4533ed02e8df1f5f784333c3e52dcae8c80f031ca2"},
+    {file = "langchain_experimental-0.0.63.tar.gz", hash = "sha256:fc894599bfac43445004a9ff60d9a28751426b2fea1979e4b2fa453c847850c4"},
 ]

 [package.dependencies]
@@ -2567,17 +2534,17 @@ langchain-core = ">=0.2.10,<0.3.0"

 [[package]]
 name = "langchain-openai"
-version = "0.1.17"
+version = "0.1.19"
 description = "An integration package connecting OpenAI and LangChain"
 optional = false
 python-versions = "<4.0,>=3.8.1"
 files = [
-    {file = "langchain_openai-0.1.17-py3-none-any.whl", hash = "sha256:30bef5574ecbbbb91b8025b2dc5a1bd81fd62157d3ad1a35d820141f31c5b443"},
-    {file = "langchain_openai-0.1.17.tar.gz", hash = "sha256:c5d70ddecdcb93e146f376bdbadbb6ec69de9ac0f402cd5b83de50b655ba85ee"},
+    {file = "langchain_openai-0.1.19-py3-none-any.whl", hash = "sha256:a7a739f1469d54cd988865420e7fc21b50fb93727b2e6da5ad30273fc61ecf19"},
+    {file = "langchain_openai-0.1.19.tar.gz", hash = "sha256:3bf342bb302d1444f4abafdf01c467dbd9b248497e1133808c4bae70396c79b3"},
 ]

 [package.dependencies]
-langchain-core = ">=0.2.20,<0.3.0"
+langchain-core = ">=0.2.24,<0.3.0"
 openai = ">=1.32.0,<2.0.0"
 tiktoken = ">=0.7,<1"

@@ -2773,23 +2740,20 @@ files = [

 [[package]]
 name = "mem0ai"
-version = "0.0.5"
+version = "0.0.9"
 description = "Long-term memory for AI Agents"
 optional = false
 python-versions = "<4.0,>=3.8"
 files = [
-    {file = "mem0ai-0.0.5-py3-none-any.whl", hash = "sha256:6f6e5356fd522adf0510322cd581476ea456fd7ccefca11b5ac050e9a6f00f36"},
-    {file = "mem0ai-0.0.5.tar.gz", hash = "sha256:f2ac35d15e4e620becb8d06b8ebeb1ffa85fac0b7cb2d3138056babec48dd5dd"},
+    {file = "mem0ai-0.0.9-py3-none-any.whl", hash = "sha256:d4de435729af4fd3d597d022ffb2af89a0630d6c3b4769792bbe27d2ce816858"},
+    {file = "mem0ai-0.0.9.tar.gz", hash = "sha256:e4374d5d04aa3f543cd3325f700e4b62f5358ae1c6fa5c44b2ff790c10c4e5f1"},
 ]

 [package.dependencies]
-boto3 = ">=1.34.144,<2.0.0"
-groq = ">=0.9.0,<0.10.0"
 openai = ">=1.33.0,<2.0.0"
 posthog = ">=3.5.0,<4.0.0"
 pydantic = ">=2.7.3,<3.0.0"
 qdrant-client = ">=1.9.1,<2.0.0"
-together = ">=1.2.1,<2.0.0"

 [[package]]
 name = "mergedeep"
@@ -3338,13 +3302,13 @@ sympy = "*"

 [[package]]
 name = "openai"
-version = "1.37.0"
+version = "1.37.1"
 description = "The official Python library for the openai API"
 optional = false
 python-versions = ">=3.7.1"
 files = [
-    {file = "openai-1.37.0-py3-none-any.whl", hash = "sha256:a903245c0ecf622f2830024acdaa78683c70abb8e9d37a497b851670864c9f73"},
-    {file = "openai-1.37.0.tar.gz", hash = "sha256:dc8197fc40ab9d431777b6620d962cc49f4544ffc3011f03ce0a805e6eb54adb"},
+    {file = "openai-1.37.1-py3-none-any.whl", hash = "sha256:9a6adda0d6ae8fce02d235c5671c399cfa40d6a281b3628914c7ebf244888ee3"},
+    {file = "openai-1.37.1.tar.gz", hash = "sha256:faf87206785a6b5d9e34555d6a3242482a6852bc802e453e2a891f68ee04ce55"},
 ]

 [package.dependencies]
@@ -3361,42 +3325,42 @@ datalib = ["numpy (>=1)", "pandas (>=1.2.3)", "pandas-stubs (>=1.1.0.11)"]

 [[package]]
 name = "opentelemetry-api"
-version = "1.25.0"
+version = "1.26.0"
 description = "OpenTelemetry Python API"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_api-1.25.0-py3-none-any.whl", hash = "sha256:757fa1aa020a0f8fa139f8959e53dec2051cc26b832e76fa839a6d76ecefd737"},
-    {file = "opentelemetry_api-1.25.0.tar.gz", hash = "sha256:77c4985f62f2614e42ce77ee4c9da5fa5f0bc1e1821085e9a47533a9323ae869"},
+    {file = "opentelemetry_api-1.26.0-py3-none-any.whl", hash = "sha256:7d7ea33adf2ceda2dd680b18b1677e4152000b37ca76e679da71ff103b943064"},
+    {file = "opentelemetry_api-1.26.0.tar.gz", hash = "sha256:2bd639e4bed5b18486fef0b5a520aaffde5a18fc225e808a1ac4df363f43a1ce"},
 ]

 [package.dependencies]
 deprecated = ">=1.2.6"
-importlib-metadata = ">=6.0,<=7.1"
+importlib-metadata = ">=6.0,<=8.0.0"

 [[package]]
 name = "opentelemetry-exporter-otlp-proto-common"
-version = "1.25.0"
+version = "1.26.0"
 description = "OpenTelemetry Protobuf encoding"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_exporter_otlp_proto_common-1.25.0-py3-none-any.whl", hash = "sha256:15637b7d580c2675f70246563363775b4e6de947871e01d0f4e3881d1848d693"},
-    {file = "opentelemetry_exporter_otlp_proto_common-1.25.0.tar.gz", hash = "sha256:c93f4e30da4eee02bacd1e004eb82ce4da143a2f8e15b987a9f603e0a85407d3"},
+    {file = "opentelemetry_exporter_otlp_proto_common-1.26.0-py3-none-any.whl", hash = "sha256:ee4d8f8891a1b9c372abf8d109409e5b81947cf66423fd998e56880057afbc71"},
+    {file = "opentelemetry_exporter_otlp_proto_common-1.26.0.tar.gz", hash = "sha256:bdbe50e2e22a1c71acaa0c8ba6efaadd58882e5a5978737a44a4c4b10d304c92"},
 ]

 [package.dependencies]
-opentelemetry-proto = "1.25.0"
+opentelemetry-proto = "1.26.0"

 [[package]]
 name = "opentelemetry-exporter-otlp-proto-grpc"
-version = "1.25.0"
+version = "1.26.0"
 description = "OpenTelemetry Collector Protobuf over gRPC Exporter"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_exporter_otlp_proto_grpc-1.25.0-py3-none-any.whl", hash = "sha256:3131028f0c0a155a64c430ca600fd658e8e37043cb13209f0109db5c1a3e4eb4"},
-    {file = "opentelemetry_exporter_otlp_proto_grpc-1.25.0.tar.gz", hash = "sha256:c0b1661415acec5af87625587efa1ccab68b873745ca0ee96b69bb1042087eac"},
+    {file = "opentelemetry_exporter_otlp_proto_grpc-1.26.0-py3-none-any.whl", hash = "sha256:e2be5eff72ebcb010675b818e8d7c2e7d61ec451755b8de67a140bc49b9b0280"},
+    {file = "opentelemetry_exporter_otlp_proto_grpc-1.26.0.tar.gz", hash = "sha256:a65b67a9a6b06ba1ec406114568e21afe88c1cdb29c464f2507d529eb906d8ae"},
 ]

 [package.dependencies]
@@ -3404,39 +3368,39 @@ deprecated = ">=1.2.6"
 googleapis-common-protos = ">=1.52,<2.0"
 grpcio = ">=1.0.0,<2.0.0"
 opentelemetry-api = ">=1.15,<2.0"
-opentelemetry-exporter-otlp-proto-common = "1.25.0"
-opentelemetry-proto = "1.25.0"
-opentelemetry-sdk = ">=1.25.0,<1.26.0"
+opentelemetry-exporter-otlp-proto-common = "1.26.0"
+opentelemetry-proto = "1.26.0"
+opentelemetry-sdk = ">=1.26.0,<1.27.0"

 [[package]]
 name = "opentelemetry-exporter-otlp-proto-http"
-version = "1.25.0"
+version = "1.26.0"
 description = "OpenTelemetry Collector Protobuf over HTTP Exporter"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_exporter_otlp_proto_http-1.25.0-py3-none-any.whl", hash = "sha256:2eca686ee11b27acd28198b3ea5e5863a53d1266b91cda47c839d95d5e0541a6"},
-    {file = "opentelemetry_exporter_otlp_proto_http-1.25.0.tar.gz", hash = "sha256:9f8723859e37c75183ea7afa73a3542f01d0fd274a5b97487ea24cb683d7d684"},
+    {file = "opentelemetry_exporter_otlp_proto_http-1.26.0-py3-none-any.whl", hash = "sha256:ee72a87c48ec977421b02f16c52ea8d884122470e0be573905237b540f4ee562"},
+    {file = "opentelemetry_exporter_otlp_proto_http-1.26.0.tar.gz", hash = "sha256:5801ebbcf7b527377883e6cbbdda35ee712dc55114fff1e93dfee210be56c908"},
 ]

 [package.dependencies]
 deprecated = ">=1.2.6"
 googleapis-common-protos = ">=1.52,<2.0"
 opentelemetry-api = ">=1.15,<2.0"
-opentelemetry-exporter-otlp-proto-common = "1.25.0"
-opentelemetry-proto = "1.25.0"
-opentelemetry-sdk = ">=1.25.0,<1.26.0"
+opentelemetry-exporter-otlp-proto-common = "1.26.0"
+opentelemetry-proto = "1.26.0"
+opentelemetry-sdk = ">=1.26.0,<1.27.0"
 requests = ">=2.7,<3.0"

 [[package]]
 name = "opentelemetry-instrumentation"
-version = "0.46b0"
+version = "0.47b0"
 description = "Instrumentation Tools & Auto Instrumentation for OpenTelemetry Python"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_instrumentation-0.46b0-py3-none-any.whl", hash = "sha256:89cd721b9c18c014ca848ccd11181e6b3fd3f6c7669e35d59c48dc527408c18b"},
-    {file = "opentelemetry_instrumentation-0.46b0.tar.gz", hash = "sha256:974e0888fb2a1e01c38fbacc9483d024bb1132aad92d6d24e2e5543887a7adda"},
+    {file = "opentelemetry_instrumentation-0.47b0-py3-none-any.whl", hash = "sha256:88974ee52b1db08fc298334b51c19d47e53099c33740e48c4f084bd1afd052d5"},
+    {file = "opentelemetry_instrumentation-0.47b0.tar.gz", hash = "sha256:96f9885e450c35e3f16a4f33145f2ebf620aea910c9fd74a392bbc0f807a350f"},
 ]

 [package.dependencies]
@@ -3446,55 +3410,55 @@ wrapt = ">=1.0.0,<2.0.0"

 [[package]]
 name = "opentelemetry-instrumentation-asgi"
-version = "0.46b0"
+version = "0.47b0"
 description = "ASGI instrumentation for OpenTelemetry"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_instrumentation_asgi-0.46b0-py3-none-any.whl", hash = "sha256:f13c55c852689573057837a9500aeeffc010c4ba59933c322e8f866573374759"},
-    {file = "opentelemetry_instrumentation_asgi-0.46b0.tar.gz", hash = "sha256:02559f30cf4b7e2a737ab17eb52aa0779bcf4cc06573064f3e2cb4dcc7d3040a"},
+    {file = "opentelemetry_instrumentation_asgi-0.47b0-py3-none-any.whl", hash = "sha256:b798dc4957b3edc9dfecb47a4c05809036a4b762234c5071212fda39ead80ade"},
+    {file = "opentelemetry_instrumentation_asgi-0.47b0.tar.gz", hash = "sha256:e78b7822c1bca0511e5e9610ec484b8994a81670375e570c76f06f69af7c506a"},
 ]

 [package.dependencies]
 asgiref = ">=3.0,<4.0"
 opentelemetry-api = ">=1.12,<2.0"
-opentelemetry-instrumentation = "0.46b0"
-opentelemetry-semantic-conventions = "0.46b0"
-opentelemetry-util-http = "0.46b0"
+opentelemetry-instrumentation = "0.47b0"
+opentelemetry-semantic-conventions = "0.47b0"
+opentelemetry-util-http = "0.47b0"

 [package.extras]
 instruments = ["asgiref (>=3.0,<4.0)"]

 [[package]]
 name = "opentelemetry-instrumentation-fastapi"
-version = "0.46b0"
+version = "0.47b0"
 description = "OpenTelemetry FastAPI Instrumentation"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_instrumentation_fastapi-0.46b0-py3-none-any.whl", hash = "sha256:e0f5d150c6c36833dd011f0e6ef5ede6d7406c1aed0c7c98b2d3b38a018d1b33"},
-    {file = "opentelemetry_instrumentation_fastapi-0.46b0.tar.gz", hash = "sha256:928a883a36fc89f9702f15edce43d1a7104da93d740281e32d50ffd03dbb4365"},
+    {file = "opentelemetry_instrumentation_fastapi-0.47b0-py3-none-any.whl", hash = "sha256:5ac28dd401160b02e4f544a85a9e4f61a8cbe5b077ea0379d411615376a2bd21"},
+    {file = "opentelemetry_instrumentation_fastapi-0.47b0.tar.gz", hash = "sha256:0c7c10b5d971e99a420678ffd16c5b1ea4f0db3b31b62faf305fbb03b4ebee36"},
 ]

 [package.dependencies]
 opentelemetry-api = ">=1.12,<2.0"
-opentelemetry-instrumentation = "0.46b0"
-opentelemetry-instrumentation-asgi = "0.46b0"
-opentelemetry-semantic-conventions = "0.46b0"
-opentelemetry-util-http = "0.46b0"
+opentelemetry-instrumentation = "0.47b0"
+opentelemetry-instrumentation-asgi = "0.47b0"
+opentelemetry-semantic-conventions = "0.47b0"
+opentelemetry-util-http = "0.47b0"

 [package.extras]
-instruments = ["fastapi (>=0.58,<1.0)"]
+instruments = ["fastapi (>=0.58,<1.0)", "fastapi-slim (>=0.111.0,<0.112.0)"]

 [[package]]
 name = "opentelemetry-proto"
-version = "1.25.0"
+version = "1.26.0"
 description = "OpenTelemetry Python Proto"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_proto-1.25.0-py3-none-any.whl", hash = "sha256:f07e3341c78d835d9b86665903b199893befa5e98866f63d22b00d0b7ca4972f"},
-    {file = "opentelemetry_proto-1.25.0.tar.gz", hash = "sha256:35b6ef9dc4a9f7853ecc5006738ad40443701e52c26099e197895cbda8b815a3"},
+    {file = "opentelemetry_proto-1.26.0-py3-none-any.whl", hash = "sha256:6c4d7b4d4d9c88543bcf8c28ae3f8f0448a753dc291c18c5390444c90b76a725"},
+    {file = "opentelemetry_proto-1.26.0.tar.gz", hash = "sha256:c5c18796c0cab3751fc3b98dee53855835e90c0422924b484432ac852d93dc1e"},
 ]

 [package.dependencies]
@@ -3502,43 +3466,44 @@ protobuf = ">=3.19,<5.0"

 [[package]]
 name = "opentelemetry-sdk"
-version = "1.25.0"
+version = "1.26.0"
 description = "OpenTelemetry Python SDK"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_sdk-1.25.0-py3-none-any.whl", hash = "sha256:d97ff7ec4b351692e9d5a15af570c693b8715ad78b8aafbec5c7100fe966b4c9"},
-    {file = "opentelemetry_sdk-1.25.0.tar.gz", hash = "sha256:ce7fc319c57707ef5bf8b74fb9f8ebdb8bfafbe11898410e0d2a761d08a98ec7"},
+    {file = "opentelemetry_sdk-1.26.0-py3-none-any.whl", hash = "sha256:feb5056a84a88670c041ea0ded9921fca559efec03905dddeb3885525e0af897"},
+    {file = "opentelemetry_sdk-1.26.0.tar.gz", hash = "sha256:c90d2868f8805619535c05562d699e2f4fb1f00dbd55a86dcefca4da6fa02f85"},
 ]

 [package.dependencies]
-opentelemetry-api = "1.25.0"
-opentelemetry-semantic-conventions = "0.46b0"
+opentelemetry-api = "1.26.0"
+opentelemetry-semantic-conventions = "0.47b0"
 typing-extensions = ">=3.7.4"

 [[package]]
 name = "opentelemetry-semantic-conventions"
-version = "0.46b0"
+version = "0.47b0"
 description = "OpenTelemetry Semantic Conventions"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_semantic_conventions-0.46b0-py3-none-any.whl", hash = "sha256:6daef4ef9fa51d51855d9f8e0ccd3a1bd59e0e545abe99ac6203804e36ab3e07"},
-    {file = "opentelemetry_semantic_conventions-0.46b0.tar.gz", hash = "sha256:fbc982ecbb6a6e90869b15c1673be90bd18c8a56ff1cffc0864e38e2edffaefa"},
+    {file = "opentelemetry_semantic_conventions-0.47b0-py3-none-any.whl", hash = "sha256:4ff9d595b85a59c1c1413f02bba320ce7ea6bf9e2ead2b0913c4395c7bbc1063"},
+    {file = "opentelemetry_semantic_conventions-0.47b0.tar.gz", hash = "sha256:a8d57999bbe3495ffd4d510de26a97dadc1dace53e0275001b2c1b2f67992a7e"},
 ]

 [package.dependencies]
-opentelemetry-api = "1.25.0"
+deprecated = ">=1.2.6"
+opentelemetry-api = "1.26.0"

 [[package]]
 name = "opentelemetry-util-http"
-version = "0.46b0"
+version = "0.47b0"
 description = "Web util for OpenTelemetry"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "opentelemetry_util_http-0.46b0-py3-none-any.whl", hash = "sha256:8dc1949ce63caef08db84ae977fdc1848fe6dc38e6bbaad0ae3e6ecd0d451629"},
-    {file = "opentelemetry_util_http-0.46b0.tar.gz", hash = "sha256:03b6e222642f9c7eae58d9132343e045b50aca9761fcb53709bd2b663571fdf6"},
+    {file = "opentelemetry_util_http-0.47b0-py3-none-any.whl", hash = "sha256:3d3215e09c4a723b12da6d0233a31395aeb2bb33a64d7b15a1500690ba250f19"},
+    {file = "opentelemetry_util_http-0.47b0.tar.gz", hash = "sha256:352a07664c18eef827eb8ddcbd64c64a7284a39dd1655e2f16f577eb046ccb32"},
 ]

 [[package]]
@@ -3917,13 +3882,13 @@ test = ["coverage", "flake8", "freezegun (==0.3.15)", "mock (>=2.0.0)", "pylint"

 [[package]]
 name = "pre-commit"
-version = "3.7.1"
+version = "3.8.0"
 description = "A framework for managing and maintaining multi-language pre-commit hooks."
 optional = false
 python-versions = ">=3.9"
 files = [
-    {file = "pre_commit-3.7.1-py2.py3-none-any.whl", hash = "sha256:fae36fd1d7ad7d6a5a1c0b0d5adb2ed1a3bda5a21bf6c3e5372073d7a11cd4c5"},
-    {file = "pre_commit-3.7.1.tar.gz", hash = "sha256:8ca3ad567bc78a4972a3f1a477e94a79d4597e8140a6e0b651c5e33899c3654a"},
+    {file = "pre_commit-3.8.0-py2.py3-none-any.whl", hash = "sha256:9a90a53bf82fdd8778d58085faf8d83df56e40dfe18f45b19446e26bf1b3a63f"},
+    {file = "pre_commit-3.8.0.tar.gz", hash = "sha256:8bb6494d4a20423842e198980c9ecf9f96607a07ea29549e180eef9ae80fe7af"},
 ]

 [package.dependencies]
@@ -3952,22 +3917,22 @@ testing = ["google-api-core (>=1.31.5)"]

 [[package]]
 name = "protobuf"
-version = "4.25.3"
+version = "4.25.4"
 description = ""
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "protobuf-4.25.3-cp310-abi3-win32.whl", hash = "sha256:d4198877797a83cbfe9bffa3803602bbe1625dc30d8a097365dbc762e5790faa"},
-    {file = "protobuf-4.25.3-cp310-abi3-win_amd64.whl", hash = "sha256:209ba4cc916bab46f64e56b85b090607a676f66b473e6b762e6f1d9d591eb2e8"},
-    {file = "protobuf-4.25.3-cp37-abi3-macosx_10_9_universal2.whl", hash = "sha256:f1279ab38ecbfae7e456a108c5c0681e4956d5b1090027c1de0f934dfdb4b35c"},
-    {file = "protobuf-4.25.3-cp37-abi3-manylinux2014_aarch64.whl", hash = "sha256:e7cb0ae90dd83727f0c0718634ed56837bfeeee29a5f82a7514c03ee1364c019"},
-    {file = "protobuf-4.25.3-cp37-abi3-manylinux2014_x86_64.whl", hash = "sha256:7c8daa26095f82482307bc717364e7c13f4f1c99659be82890dcfc215194554d"},
-    {file = "protobuf-4.25.3-cp38-cp38-win32.whl", hash = "sha256:f4f118245c4a087776e0a8408be33cf09f6c547442c00395fbfb116fac2f8ac2"},
-    {file = "protobuf-4.25.3-cp38-cp38-win_amd64.whl", hash = "sha256:c053062984e61144385022e53678fbded7aea14ebb3e0305ae3592fb219ccfa4"},
-    {file = "protobuf-4.25.3-cp39-cp39-win32.whl", hash = "sha256:19b270aeaa0099f16d3ca02628546b8baefe2955bbe23224aaf856134eccf1e4"},
-    {file = "protobuf-4.25.3-cp39-cp39-win_amd64.whl", hash = "sha256:e3c97a1555fd6388f857770ff8b9703083de6bf1f9274a002a332d65fbb56c8c"},
-    {file = "protobuf-4.25.3-py3-none-any.whl", hash = "sha256:f0700d54bcf45424477e46a9f0944155b46fb0639d69728739c0e47bab83f2b9"},
-    {file = "protobuf-4.25.3.tar.gz", hash = "sha256:25b5d0b42fd000320bd7830b349e3b696435f3b329810427a6bcce6a5492cc5c"},
+    {file = "protobuf-4.25.4-cp310-abi3-win32.whl", hash = "sha256:db9fd45183e1a67722cafa5c1da3e85c6492a5383f127c86c4c4aa4845867dc4"},
+    {file = "protobuf-4.25.4-cp310-abi3-win_amd64.whl", hash = "sha256:ba3d8504116a921af46499471c63a85260c1a5fc23333154a427a310e015d26d"},
+    {file = "protobuf-4.25.4-cp37-abi3-macosx_10_9_universal2.whl", hash = "sha256:eecd41bfc0e4b1bd3fa7909ed93dd14dd5567b98c941d6c1ad08fdcab3d6884b"},
+    {file = "protobuf-4.25.4-cp37-abi3-manylinux2014_aarch64.whl", hash = "sha256:4c8a70fdcb995dcf6c8966cfa3a29101916f7225e9afe3ced4395359955d3835"},
+    {file = "protobuf-4.25.4-cp37-abi3-manylinux2014_x86_64.whl", hash = "sha256:3319e073562e2515c6ddc643eb92ce20809f5d8f10fead3332f71c63be6a7040"},
+    {file = "protobuf-4.25.4-cp38-cp38-win32.whl", hash = "sha256:7e372cbbda66a63ebca18f8ffaa6948455dfecc4e9c1029312f6c2edcd86c4e1"},
+    {file = "protobuf-4.25.4-cp38-cp38-win_amd64.whl", hash = "sha256:051e97ce9fa6067a4546e75cb14f90cf0232dcb3e3d508c448b8d0e4265b61c1"},
+    {file = "protobuf-4.25.4-cp39-cp39-win32.whl", hash = "sha256:90bf6fd378494eb698805bbbe7afe6c5d12c8e17fca817a646cd6a1818c696ca"},
+    {file = "protobuf-4.25.4-cp39-cp39-win_amd64.whl", hash = "sha256:ac79a48d6b99dfed2729ccccee547b34a1d3d63289c71cef056653a846a2240f"},
+    {file = "protobuf-4.25.4-py3-none-any.whl", hash = "sha256:bfbebc1c8e4793cfd58589acfb8a1026be0003e852b9da7db5a4285bde996978"},
+    {file = "protobuf-4.25.4.tar.gz", hash = "sha256:0dc4a62cc4052a036ee2204d26fe4d835c62827c855c8a03f29fe6da146b380d"},
 ]

 [[package]]
@@ -4317,13 +4282,13 @@ torch = ["torch"]

 [[package]]
 name = "pymdown-extensions"
-version = "10.8.1"
+version = "10.9"
 description = "Extension pack for Python Markdown."
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "pymdown_extensions-10.8.1-py3-none-any.whl", hash = "sha256:f938326115884f48c6059c67377c46cf631c733ef3629b6eed1349989d1b30cb"},
-    {file = "pymdown_extensions-10.8.1.tar.gz", hash = "sha256:3ab1db5c9e21728dabf75192d71471f8e50f216627e9a1fa9535ecb0231b9940"},
+    {file = "pymdown_extensions-10.9-py3-none-any.whl", hash = "sha256:d323f7e90d83c86113ee78f3fe62fc9dee5f56b54d912660703ea1816fed5626"},
+    {file = "pymdown_extensions-10.9.tar.gz", hash = "sha256:6ff740bcd99ec4172a938970d42b96128bdc9d4b9bcad72494f29921dc69b753"},
 ]

 [package.dependencies]
@@ -4388,13 +4353,13 @@ files = [

 [[package]]
 name = "pyright"
-version = "1.1.372"
+version = "1.1.373"
 description = "Command line wrapper for pyright"
 optional = false
 python-versions = ">=3.7"
 files = [
-    {file = "pyright-1.1.372-py3-none-any.whl", hash = "sha256:25b15fb8967740f0949fd35b963777187f0a0404c0bd753cc966ec139f3eaa0b"},
-    {file = "pyright-1.1.372.tar.gz", hash = "sha256:a9f5e0daa955daaa17e3d1ef76d3623e75f8afd5e37b437d3ff84d5b38c15420"},
+    {file = "pyright-1.1.373-py3-none-any.whl", hash = "sha256:b805413227f2c209f27b14b55da27fe5e9fb84129c9f1eb27708a5d12f6f000e"},
+    {file = "pyright-1.1.373.tar.gz", hash = "sha256:f41bcfc8b9d1802b09921a394d6ae1ce19694957b628bc657629688daf8a83ff"},
 ]

 [package.dependencies]
@@ -4428,13 +4393,13 @@ files = [

 [[package]]
 name = "pytest"
-version = "8.3.1"
+version = "8.3.2"
 description = "pytest: simple powerful testing with Python"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "pytest-8.3.1-py3-none-any.whl", hash = "sha256:e9600ccf4f563976e2c99fa02c7624ab938296551f280835ee6516df8bc4ae8c"},
-    {file = "pytest-8.3.1.tar.gz", hash = "sha256:7e8e5c5abd6e93cb1cc151f23e57adc31fcf8cfd2a3ff2da63e23f732de35db6"},
+    {file = "pytest-8.3.2-py3-none-any.whl", hash = "sha256:4ba08f9ae7dcf84ded419494d229b48d0903ea6407b030eaec46df5e6a73bba5"},
+    {file = "pytest-8.3.2.tar.gz", hash = "sha256:c132345d12ce551242c87269de812483f5bcc87cdbb4722e48487ba194f9fdce"},
 ]

 [package.dependencies]
@@ -4448,6 +4413,24 @@ tomli = {version = ">=1", markers = "python_version < \"3.11\""}
 [package.extras]
 dev = ["argcomplete", "attrs (>=19.2)", "hypothesis (>=3.56)", "mock", "pygments (>=2.7.2)", "requests", "setuptools", "xmlschema"]

+[[package]]
+name = "pytest-asyncio"
+version = "0.23.8"
+description = "Pytest support for asyncio"
+optional = false
+python-versions = ">=3.8"
+files = [
+    {file = "pytest_asyncio-0.23.8-py3-none-any.whl", hash = "sha256:50265d892689a5faefb84df80819d1ecef566eb3549cf915dfb33569359d1ce2"},
+    {file = "pytest_asyncio-0.23.8.tar.gz", hash = "sha256:759b10b33a6dc61cce40a8bd5205e302978bbbcc00e279a8b61d9a6a3c82e4d3"},
+]
+
+[package.dependencies]
+pytest = ">=7.0.0,<9"
+
+[package.extras]
+docs = ["sphinx (>=5.3)", "sphinx-rtd-theme (>=1.0)"]
+testing = ["coverage (>=6.2)", "hypothesis (>=5.7.1)"]
+
 [[package]]
 name = "pytest-vcr"
 version = "1.0.2"
@@ -4883,22 +4866,22 @@ files = [

 [[package]]
 name = "selenium"
-version = "4.22.0"
+version = "4.23.1"
 description = "Official Python bindings for Selenium WebDriver"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "selenium-4.22.0-py3-none-any.whl", hash = "sha256:e424991196e9857e19bf04fe5c1c0a4aac076794ff5e74615b1124e729d93104"},
-    {file = "selenium-4.22.0.tar.gz", hash = "sha256:903c8c9d61b3eea6fcc9809dc7d9377e04e2ac87709876542cc8f863e482c4ce"},
+    {file = "selenium-4.23.1-py3-none-any.whl", hash = "sha256:3a8d9f23dc636bd3840dd56f00c2739e32ec0c1e34a821dd553e15babef24477"},
+    {file = "selenium-4.23.1.tar.gz", hash = "sha256:128d099e66284437e7128d2279176ec7a06e6ec7426e167f5d34987166bd8f46"},
 ]

 [package.dependencies]
 certifi = ">=2021.10.8"
 trio = ">=0.17,<1.0"
 trio-websocket = ">=0.9,<1.0"
-typing_extensions = ">=4.9.0"
+typing_extensions = ">=4.9,<5.0"
 urllib3 = {version = ">=1.26,<3", extras = ["socks"]}
-websocket-client = ">=1.8.0"
+websocket-client = ">=1.8,<2.0"

 [[package]]
 name = "semver"
@@ -4913,13 +4896,13 @@ files = [

 [[package]]
 name = "setuptools"
-version = "71.1.0"
+version = "72.1.0"
 description = "Easily download, build, install, upgrade, and uninstall Python packages"
 optional = false
 python-versions = ">=3.8"
 files = [
-    {file = "setuptools-71.1.0-py3-none-any.whl", hash = "sha256:33874fdc59b3188304b2e7c80d9029097ea31627180896fb549c578ceb8a0855"},
-    {file = "setuptools-71.1.0.tar.gz", hash = "sha256:032d42ee9fb536e33087fb66cac5f840eb9391ed05637b3f2a76a7c8fb477936"},
+    {file = "setuptools-72.1.0-py3-none-any.whl", hash = "sha256:5a03e1860cf56bb6ef48ce186b0e557fdba433237481a9a625176c2831be15d1"},
+    {file = "setuptools-72.1.0.tar.gz", hash = "sha256:8d243eff56d095e5817f796ede6ae32941278f542e0f941867cc05ae52b162ec"},
 ]

 [package.extras]
@@ -5268,34 +5251,6 @@ webencodings = ">=0.4"
 doc = ["sphinx", "sphinx_rtd_theme"]
 test = ["pytest", "ruff"]

-[[package]]
-name = "together"
-version = "1.2.3"
-description = "Python client for Together's Cloud Platform!"
-optional = false
-python-versions = "<4.0,>=3.8"
-files = [
-    {file = "together-1.2.3-py3-none-any.whl", hash = "sha256:bbafb4b8340e0f7e0ddb11ad447eb3467c591090910d0291cfbf74b47af045c1"},
-    {file = "together-1.2.3.tar.gz", hash = "sha256:4ea7626a9581d16fbf293e3eaf91557c43dea044627cf6dbe458bbf43408a6b2"},
-]
-
-[package.dependencies]
-aiohttp = ">=3.9.3,<4.0.0"
-click = ">=8.1.7,<9.0.0"
-eval-type-backport = ">=0.1.3,<0.3.0"
-filelock = ">=3.13.1,<4.0.0"
-numpy = [
-    {version = ">=1.23.5", markers = "python_version < \"3.12\""},
-    {version = ">=1.26.0", markers = "python_version >= \"3.12\""},
-]
-pillow = ">=10.3.0,<11.0.0"
-pyarrow = ">=10.0.1"
-pydantic = ">=2.6.3,<3.0.0"
-requests = ">=2.31.0,<3.0.0"
-tabulate = ">=0.9.0,<0.10.0"
-tqdm = ">=4.66.2,<5.0.0"
-typer = ">=0.9,<0.13"
-
 [[package]]
 name = "tokenizers"
 version = "0.19.1"
@@ -6144,4 +6099,4 @@ tools = ["crewai-tools"]
 [metadata]
 lock-version = "2.0"
 python-versions = ">=3.10,<=3.13"
-content-hash = "f5ad9babb3c57c405e39232020e8cbfaaeb5c315c2e7c5bb8fdf66792f260343"
+content-hash = "8df022f5ec0997c0a0f5710476139d9117c1057889c158e958f2c8efd22a4756"
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -52,6 +52,7 @@ crewai-tools = "^0.4.26"
 pytest = "^8.0.0"
 pytest-vcr = "^1.0.2"
 python-dotenv = "1.0.0"
+pytest-asyncio = "^0.23.7"

 [tool.poetry.scripts]
 crewai = "crewai.cli.cli:crewai"
--- a/src/crewai/agents/agent_builder/utilities/base_token_process.py
+++ b/src/crewai/agents/agent_builder/utilities/base_token_process.py
@@ -1,4 +1,4 @@
-from typing import Any, Dict
+from crewai.types.usage_metrics import UsageMetrics


 class TokenProcess:
@@ -18,10 +18,10 @@ class TokenProcess:
    def sum_successful_requests(self, requests: int):
        self.successful_requests = self.successful_requests + requests

-    def get_summary(self) -> Dict[str, Any]:
-        return {
-            "total_tokens": self.total_tokens,
-            "prompt_tokens": self.prompt_tokens,
-            "completion_tokens": self.completion_tokens,
-            "successful_requests": self.successful_requests,
-        }
+    def get_summary(self) -> UsageMetrics:
+        return UsageMetrics(
+            total_tokens=self.total_tokens,
+            prompt_tokens=self.prompt_tokens,
+            completion_tokens=self.completion_tokens,
+            successful_requests=self.successful_requests,
+        )
--- a/src/crewai/agents/executor.py
+++ b/src/crewai/agents/executor.py
@@ -56,7 +56,7 @@ class CrewAgentExecutor(AgentExecutor, CrewAgentExecutorMixin):
        )
        intermediate_steps: List[Tuple[AgentAction, str]] = []
        # Allowing human input given task setting
-        if self.task.human_input:
+        if self.task and self.task.human_input:
            self.should_ask_for_human_input = True

        # Let's start tracking the number of iterations and time elapsed
--- a/src/crewai/crew.py
+++ b/src/crewai/crew.py
@@ -3,7 +3,7 @@ import json
 import uuid
 from concurrent.futures import Future
 from hashlib import md5
-from typing import Any, Dict, List, Optional, Tuple, Union
+from typing import TYPE_CHECKING, Any, Dict, List, Optional, Tuple, Union

 from langchain_core.callbacks import BaseCallbackHandler
 from pydantic import (
@@ -32,11 +32,9 @@ from crewai.tasks.conditional_task import ConditionalTask
 from crewai.tasks.task_output import TaskOutput
 from crewai.telemetry import Telemetry
 from crewai.tools.agent_tools import AgentTools
+from crewai.types.usage_metrics import UsageMetrics
 from crewai.utilities import I18N, FileHandler, Logger, RPMController
-from crewai.utilities.constants import (
-    TRAINED_AGENTS_DATA_FILE,
-    TRAINING_DATA_FILE,
-)
+from crewai.utilities.constants import TRAINED_AGENTS_DATA_FILE, TRAINING_DATA_FILE
 from crewai.utilities.evaluators.crew_evaluator_handler import CrewEvaluator
 from crewai.utilities.evaluators.task_evaluator import TaskEvaluator
 from crewai.utilities.formatter import (
@@ -52,6 +50,9 @@ try:
 except ImportError:
    agentops = None

+if TYPE_CHECKING:
+    from crewai.pipeline.pipeline import Pipeline
+

 class Crew(BaseModel):
    """
@@ -97,12 +98,13 @@ class Crew(BaseModel):
        default_factory=TaskOutputStorageHandler
    )

+    name: Optional[str] = Field(default=None)
    cache: bool = Field(default=True)
    model_config = ConfigDict(arbitrary_types_allowed=True)
    tasks: List[Task] = Field(default_factory=list)
    agents: List[BaseAgent] = Field(default_factory=list)
    process: Process = Field(default=Process.sequential)
-    verbose: Union[int, bool] = Field(default=0)
+    verbose: int = Field(default=0)
    memory: bool = Field(
        default=False,
        description="Whether the crew should use memory to store memories of it's execution",
@@ -111,7 +113,7 @@ class Crew(BaseModel):
        default={"provider": "openai"},
        description="Configuration for the embedder to be used for the crew.",
    )
-    usage_metrics: Optional[dict] = Field(
+    usage_metrics: Optional[UsageMetrics] = Field(
        default=None,
        description="Metrics for the LLM usage during all tasks execution.",
    )
@@ -147,8 +149,8 @@ class Crew(BaseModel):
        default=None,
        description="Path to the prompt json file to be used for the crew.",
    )
-    output_log_file: Optional[Union[bool, str]] = Field(
-        default=False,
+    output_log_file: Optional[str] = Field(
+        default=None,
        description="output_log_file",
    )
    planning: Optional[bool] = Field(
@@ -453,7 +455,7 @@ class Crew(BaseModel):
        if self.planning:
            self._handle_crew_planning()

-        metrics = []
+        metrics: List[UsageMetrics] = []

        if self.process == Process.sequential:
            result = self._run_sequential_process()
@@ -463,11 +465,12 @@ class Crew(BaseModel):
            raise NotImplementedError(
                f"The process '{self.process}' is not implemented yet."
            )
+
        metrics += [agent._token_process.get_summary() for agent in self.agents]

-        self.usage_metrics = {
-            key: sum([m[key] for m in metrics if m is not None]) for key in metrics[0]
-        }
+        self.usage_metrics = UsageMetrics()
+        for metric in metrics:
+            self.usage_metrics.add_usage_metrics(metric)

        return result

@@ -476,12 +479,7 @@ class Crew(BaseModel):
        results: List[CrewOutput] = []

        # Initialize the parent crew's usage metrics
-        total_usage_metrics = {
-            "total_tokens": 0,
-            "prompt_tokens": 0,
-            "completion_tokens": 0,
-            "successful_requests": 0,
-        }
+        total_usage_metrics = UsageMetrics()

        for input_data in inputs:
            crew = self.copy()
@@ -489,8 +487,7 @@ class Crew(BaseModel):
            output = crew.kickoff(inputs=input_data)

            if crew.usage_metrics:
-                for key in total_usage_metrics:
-                    total_usage_metrics[key] += crew.usage_metrics.get(key, 0)
+                total_usage_metrics.add_usage_metrics(crew.usage_metrics)

            results.append(output)

@@ -519,29 +516,10 @@ class Crew(BaseModel):

        results = await asyncio.gather(*tasks)

-        total_usage_metrics = {
-            "total_tokens": 0,
-            "prompt_tokens": 0,
-            "completion_tokens": 0,
-            "successful_requests": 0,
-        }
+        total_usage_metrics = UsageMetrics()
        for crew in crew_copies:
            if crew.usage_metrics:
-                for key in total_usage_metrics:
-                    total_usage_metrics[key] += crew.usage_metrics.get(key, 0)
-
-        self.usage_metrics = total_usage_metrics
-
-        total_usage_metrics = {
-            "total_tokens": 0,
-            "prompt_tokens": 0,
-            "completion_tokens": 0,
-            "successful_requests": 0,
-        }
-        for crew in crew_copies:
-            if crew.usage_metrics:
-                for key in total_usage_metrics:
-                    total_usage_metrics[key] += crew.usage_metrics.get(key, 0)
+                total_usage_metrics.add_usage_metrics(crew.usage_metrics)

        self.usage_metrics = total_usage_metrics
        self._task_output_handler.reset()
@@ -932,25 +910,18 @@ class Crew(BaseModel):
            )
        self._telemetry.end_crew(self, final_string_output)

-    def calculate_usage_metrics(self) -> Dict[str, int]:
+    def calculate_usage_metrics(self) -> UsageMetrics:
        """Calculates and returns the usage metrics."""
-        total_usage_metrics = {
-            "total_tokens": 0,
-            "prompt_tokens": 0,
-            "completion_tokens": 0,
-            "successful_requests": 0,
-        }
+        total_usage_metrics = UsageMetrics()

        for agent in self.agents:
            if hasattr(agent, "_token_process"):
                token_sum = agent._token_process.get_summary()
-                for key in total_usage_metrics:
-                    total_usage_metrics[key] += token_sum.get(key, 0)
+                total_usage_metrics.add_usage_metrics(token_sum)

        if self.manager_agent and hasattr(self.manager_agent, "_token_process"):
            token_sum = self.manager_agent._token_process.get_summary()
-            for key in total_usage_metrics:
-                total_usage_metrics[key] += token_sum.get(key, 0)
+            total_usage_metrics.add_usage_metrics(token_sum)

        return total_usage_metrics

@@ -969,5 +940,17 @@ class Crew(BaseModel):

        evaluator.print_crew_evaluation_result()

+    def __rshift__(self, other: "Crew") -> "Pipeline":
+        """
+        Implements the >> operator to add another Crew to an existing Pipeline.
+        """
+        from crewai.pipeline.pipeline import Pipeline
+
+        if not isinstance(other, Crew):
+            raise TypeError(
+                f"Unsupported operand type for >>: '{type(self).__name__}' and '{type(other).__name__}'"
+            )
+        return Pipeline(stages=[self, other])
+
    def __repr__(self):
        return f"Crew(id={self.id}, process={self.process}, number_of_agents={len(self.agents)}, number_of_tasks={len(self.tasks)})"
--- a/src/crewai/crews/crew_output.py
+++ b/src/crewai/crews/crew_output.py
@@ -5,6 +5,7 @@ from pydantic import BaseModel, Field

 from crewai.tasks.output_format import OutputFormat
 from crewai.tasks.task_output import TaskOutput
+from crewai.types.usage_metrics import UsageMetrics


 class CrewOutput(BaseModel):
@@ -20,9 +21,7 @@ class CrewOutput(BaseModel):
    tasks_output: list[TaskOutput] = Field(
        description="Output of each task", default=[]
    )
-    token_usage: Dict[str, Any] = Field(
-        description="Processed token summary", default={}
-    )
+    token_usage: UsageMetrics = Field(description="Processed token summary", default={})

    @property
    def json(self) -> Optional[str]:
--- a/src/crewai/memory/storage/rag_storage.py
+++ b/src/crewai/memory/storage/rag_storage.py
@@ -5,12 +5,12 @@ import os
 import shutil
 from typing import Any, Dict, List, Optional

-from embedchain import App
-from embedchain.llm.base import BaseLlm
-from embedchain.vectordb.chroma import InvalidDimensionException
-
 from crewai.memory.storage.interface import Storage
 from crewai.utilities.paths import db_storage_path
+from embedchain import App
+from embedchain.llm.base import BaseLlm
+from embedchain.models.data_type import DataType
+from embedchain.vectordb.chroma import InvalidDimensionException


@contextlib.contextmanager
@@ -101,8 +101,7 @@ class RAGStorage(Storage):
        return [r for r in results if r["metadata"]["score"] >= score_threshold]

    def _generate_embedding(self, text: str, metadata: Dict[str, Any]) -> Any:
-        with suppress_logging():
-            self.app.add(text, data_type="text", metadata=metadata)
+        self.app.add(text, data_type=DataType.TEXT, metadata=metadata)

    def reset(self) -> None:
        try:
--- a/src/crewai/pipeline/init.py
+++ b/src/crewai/pipeline/init.py
@@ -0,0 +1,3 @@
+from crewai.pipeline.pipeline import Pipeline
+from crewai.pipeline.pipeline_kickoff_result import PipelineKickoffResult
+from crewai.pipeline.pipeline_output import PipelineOutput
--- a/src/crewai/pipeline/pipeline.py
+++ b/src/crewai/pipeline/pipeline.py
@@ -0,0 +1,387 @@
+import asyncio
+import copy
+from typing import Any, Dict, List, Tuple, Union
+
+from pydantic import BaseModel, Field, model_validator
+
+from crewai.crew import Crew
+from crewai.crews.crew_output import CrewOutput
+from crewai.pipeline.pipeline_kickoff_result import PipelineKickoffResult
+from crewai.routers.router import Router
+from crewai.types.usage_metrics import UsageMetrics
+
+Trace = Union[Union[str, Dict[str, Any]], List[Union[str, Dict[str, Any]]]]
+PipelineStage = Union[Crew, List[Crew], Router]
+
+"""
+Developer Notes:
+
+This module defines a Pipeline class that represents a sequence of operations (stages)
+to process inputs. Each stage can be either sequential or parallel, and the pipeline
+can process multiple kickoffs concurrently.
+
+Core Loop Explanation:
+1. The `process_kickoffs` method processes multiple kickoffs in parallel, each going through
+   all pipeline stages.
+2. The `process_single_kickoff` method handles the processing of a single kickouff through
+   all stages, updating metrics and input data along the way.
+3. The `_process_stage` method determines whether a stage is sequential or parallel
+   and processes it accordingly.
+4. The `_process_single_crew` and `_process_parallel_crews` methods handle the
+   execution of single and parallel crew stages.
+5. The `_update_metrics_and_input` method updates usage metrics and the current input
+   with the outputs from a stage.
+6. The `_build_pipeline_kickoff_results` method constructs the final results of the
+   pipeline kickoff, including traces and outputs.
+
+Handling Traces and Crew Outputs:
+- During the processing of stages, we handle the results (traces and crew outputs)
+  for all stages except the last one differently from the final stage.
+- For intermediate stages, the primary focus is on passing the input data between stages.
+  This involves merging the output dictionaries from all crews in a stage into a single
+  dictionary and passing it to the next stage. This merged dictionary allows for smooth
+  data flow between stages.
+- For the final stage, in addition to passing the input data, we also need to prepare
+  the final outputs and traces to be returned as the overall result of the pipeline kickoff.
+  In this case, we do not merge the results, as each result needs to be included
+  separately in its own pipeline kickoff result.
+
+Pipeline Terminology:
+- Pipeline: The overall structure that defines a sequence of operations.
+- Stage: A distinct part of the pipeline, which can be either sequential or parallel.
+- Kickoff: A specific execution of the pipeline for a given set of inputs, representing a single instance of processing through the pipeline.
+- Branch: Parallel executions within a stage (e.g., concurrent crew operations).
+- Trace: The journey of an individual input through the entire pipeline.
+
+Example pipeline structure:
+crew1 >> crew2 >> crew3 
+
+This represents a pipeline with three sequential stages:
+1. crew1 is the first stage, which processes the input and passes its output to crew2.
+2. crew2 is the second stage, which takes the output from crew1 as its input, processes it, and passes its output to crew3.
+3. crew3 is the final stage, which takes the output from crew2 as its input and produces the final output of the pipeline.
+
+Each input creates its own kickoff, flowing through all stages of the pipeline.
+Multiple kickoffss can be processed concurrently, each following the defined pipeline structure.
+
+Another example pipeline structure:
+crew1 >> [crew2, crew3] >> crew4
+
+This represents a pipeline with three stages:
+1. A sequential stage (crew1)
+2. A parallel stage with two branches (crew2 and crew3 executing concurrently)
+3. Another sequential stage (crew4)
+
+Each input creates its own kickoff, flowing through all stages of the pipeline.
+Multiple kickoffs can be processed concurrently, each following the defined pipeline structure.
+"""
+
+
+class Pipeline(BaseModel):
+    stages: List[PipelineStage] = Field(
+        ..., description="List of crews representing stages to be executed in sequence"
+    )
+
+    @model_validator(mode="before")
+    @classmethod
+    def validate_stages(cls, values):
+        stages = values.get("stages", [])
+
+        def check_nesting_and_type(item, depth=0):
+            if depth > 1:
+                raise ValueError("Double nesting is not allowed in pipeline stages")
+            if isinstance(item, list):
+                for sub_item in item:
+                    check_nesting_and_type(sub_item, depth + 1)
+            elif not isinstance(item, (Crew, Router)):
+                raise ValueError(
+                    f"Expected Crew instance, Router instance, or list of Crews, got {type(item)}"
+                )
+
+        for stage in stages:
+            check_nesting_and_type(stage)
+        return values
+
+    async def kickoff(
+        self, inputs: List[Dict[str, Any]]
+    ) -> List[PipelineKickoffResult]:
+        """
+        Processes multiple runs in parallel, each going through all pipeline stages.
+
+        Args:
+            inputs (List[Dict[str, Any]]): List of inputs for each run.
+
+        Returns:
+            List[PipelineKickoffResult]: List of results from each run.
+        """
+        pipeline_results: List[PipelineKickoffResult] = []
+
+        # Process all runs in parallel
+        all_run_results = await asyncio.gather(
+            *(self.process_single_kickoff(input_data) for input_data in inputs)
+        )
+
+        # Flatten the list of lists into a single list of results
+        pipeline_results.extend(
+            result for run_result in all_run_results for result in run_result
+        )
+
+        return pipeline_results
+
+    async def process_single_kickoff(
+        self, kickoff_input: Dict[str, Any]
+    ) -> List[PipelineKickoffResult]:
+        """
+        Processes a single run through all pipeline stages.
+
+        Args:
+            input (Dict[str, Any]): The input for the run.
+
+        Returns:
+            List[PipelineKickoffResult]: The results of processing the run.
+        """
+        initial_input = copy.deepcopy(kickoff_input)
+        current_input = copy.deepcopy(kickoff_input)
+        stages = copy.deepcopy(self.stages)
+        pipeline_usage_metrics: Dict[str, UsageMetrics] = {}
+        all_stage_outputs: List[List[CrewOutput]] = []
+        traces: List[List[Union[str, Dict[str, Any]]]] = [[initial_input]]
+
+        stage_index = 0
+        while stage_index < len(stages):
+            stage = stages[stage_index]
+            stage_input = copy.deepcopy(current_input)
+
+            if isinstance(stage, Router):
+                next_pipeline, route_taken = stage.route(stage_input)
+                stages = (
+                    stages[: stage_index + 1]
+                    + list(next_pipeline.stages)
+                    + stages[stage_index + 1 :]
+                )
+                traces.append([{"route_taken": route_taken}])
+                stage_index += 1
+                continue
+
+            stage_outputs, stage_trace = await self._process_stage(stage, stage_input)
+
+            self._update_metrics_and_input(
+                pipeline_usage_metrics, current_input, stage, stage_outputs
+            )
+            traces.append(stage_trace)
+            all_stage_outputs.append(stage_outputs)
+            stage_index += 1
+
+        return self._build_pipeline_kickoff_results(
+            all_stage_outputs, traces, pipeline_usage_metrics
+        )
+
+    async def _process_stage(
+        self, stage: PipelineStage, current_input: Dict[str, Any]
+    ) -> Tuple[List[CrewOutput], List[Union[str, Dict[str, Any]]]]:
+        """
+        Processes a single stage of the pipeline, which can be either sequential or parallel.
+
+        Args:
+            stage (Union[Crew, List[Crew]]): The stage to process.
+            current_input (Dict[str, Any]): The input for the stage.
+
+        Returns:
+            Tuple[List[CrewOutput], List[Union[str, Dict[str, Any]]]]: The outputs and trace of the stage.
+        """
+        if isinstance(stage, Crew):
+            return await self._process_single_crew(stage, current_input)
+        elif isinstance(stage, list) and all(isinstance(crew, Crew) for crew in stage):
+            return await self._process_parallel_crews(stage, current_input)
+        else:
+            raise ValueError(f"Unsupported stage type: {type(stage)}")
+
+    async def _process_single_crew(
+        self, crew: Crew, current_input: Dict[str, Any]
+    ) -> Tuple[List[CrewOutput], List[Union[str, Dict[str, Any]]]]:
+        """
+        Processes a single crew.
+
+        Args:
+            crew (Crew): The crew to process.
+            current_input (Dict[str, Any]): The input for the crew.
+
+        Returns:
+            Tuple[List[CrewOutput], List[Union[str, Dict[str, Any]]]]: The output and trace of the crew.
+        """
+        output = await crew.kickoff_async(inputs=current_input)
+        return [output], [crew.name or str(crew.id)]
+
+    async def _process_parallel_crews(
+        self, crews: List[Crew], current_input: Dict[str, Any]
+    ) -> Tuple[List[CrewOutput], List[Union[str, Dict[str, Any]]]]:
+        """
+        Processes multiple crews in parallel.
+
+        Args:
+            crews (List[Crew]): The list of crews to process in parallel.
+            current_input (Dict[str, Any]): The input for the crews.
+
+        Returns:
+            Tuple[List[CrewOutput], List[Union[str, Dict[str, Any]]]]: The outputs and traces of the crews.
+        """
+        parallel_outputs = await asyncio.gather(
+            *[crew.kickoff_async(inputs=current_input) for crew in crews]
+        )
+        return parallel_outputs, [crew.name or str(crew.id) for crew in crews]
+
+    def _update_metrics_and_input(
+        self,
+        usage_metrics: Dict[str, UsageMetrics],
+        current_input: Dict[str, Any],
+        stage: PipelineStage,
+        outputs: List[CrewOutput],
+    ) -> None:
+        """
+        Updates metrics and current input with the outputs of a stage.
+
+        Args:
+            usage_metrics (Dict[str, Any]): The usage metrics to update.
+            current_input (Dict[str, Any]): The current input to update.
+            stage (Union[Crew, List[Crew]]): The stage that was processed.
+            outputs (List[CrewOutput]): The outputs of the stage.
+        """
+        if isinstance(stage, Crew):
+            usage_metrics[stage.name or str(stage.id)] = outputs[0].token_usage
+            current_input.update(outputs[0].to_dict())
+        elif isinstance(stage, list) and all(isinstance(crew, Crew) for crew in stage):
+            for crew, output in zip(stage, outputs):
+                usage_metrics[crew.name or str(crew.id)] = output.token_usage
+                current_input.update(output.to_dict())
+        else:
+            raise ValueError(f"Unsupported stage type: {type(stage)}")
+
+    def _build_pipeline_kickoff_results(
+        self,
+        all_stage_outputs: List[List[CrewOutput]],
+        traces: List[List[Union[str, Dict[str, Any]]]],
+        token_usage: Dict[str, UsageMetrics],
+    ) -> List[PipelineKickoffResult]:
+        """
+        Builds the results of a pipeline run.
+
+        Args:
+            all_stage_outputs (List[List[CrewOutput]]): All stage outputs.
+            traces (List[List[Union[str, Dict[str, Any]]]]): All traces.
+            token_usage (Dict[str, Any]): Token usage metrics.
+
+        Returns:
+            List[PipelineKickoffResult]: The results of the pipeline run.
+        """
+        formatted_traces = self._format_traces(traces)
+        formatted_crew_outputs = self._format_crew_outputs(all_stage_outputs)
+
+        return [
+            PipelineKickoffResult(
+                token_usage=token_usage,
+                trace=formatted_trace,
+                raw=crews_outputs[-1].raw,
+                pydantic=crews_outputs[-1].pydantic,
+                json_dict=crews_outputs[-1].json_dict,
+                crews_outputs=crews_outputs,
+            )
+            for crews_outputs, formatted_trace in zip(
+                formatted_crew_outputs, formatted_traces
+            )
+        ]
+
+    def _format_traces(
+        self, traces: List[List[Union[str, Dict[str, Any]]]]
+    ) -> List[List[Trace]]:
+        """
+        Formats the traces of a pipeline run.
+
+        Args:
+            traces (List[List[Union[str, Dict[str, Any]]]]): The traces to format.
+
+        Returns:
+            List[List[Trace]]: The formatted traces.
+        """
+        formatted_traces: List[Trace] = self._format_single_trace(traces[:-1])
+        return self._format_multiple_traces(formatted_traces, traces[-1])
+
+    def _format_single_trace(
+        self, traces: List[List[Union[str, Dict[str, Any]]]]
+    ) -> List[Trace]:
+        """
+        Formats single traces.
+
+        Args:
+            traces (List[List[Union[str, Dict[str, Any]]]]): The traces to format.
+
+        Returns:
+            List[Trace]: The formatted single traces.
+        """
+        formatted_traces: List[Trace] = []
+        for trace in traces:
+            formatted_traces.append(trace[0] if len(trace) == 1 else trace)
+        return formatted_traces
+
+    def _format_multiple_traces(
+        self,
+        formatted_traces: List[Trace],
+        final_trace: List[Union[str, Dict[str, Any]]],
+    ) -> List[List[Trace]]:
+        """
+        Formats multiple traces.
+
+        Args:
+            formatted_traces (List[Trace]): The formatted single traces.
+            final_trace (List[Union[str, Dict[str, Any]]]): The final trace to format.
+
+        Returns:
+            List[List[Trace]]: The formatted multiple traces.
+        """
+        traces_to_return: List[List[Trace]] = []
+        if len(final_trace) == 1:
+            formatted_traces.append(final_trace[0])
+            traces_to_return.append(formatted_traces)
+        else:
+            for trace in final_trace:
+                copied_traces = formatted_traces.copy()
+                copied_traces.append(trace)
+                traces_to_return.append(copied_traces)
+        return traces_to_return
+
+    def _format_crew_outputs(
+        self, all_stage_outputs: List[List[CrewOutput]]
+    ) -> List[List[CrewOutput]]:
+        """
+        Formats the outputs of all stages into a list of crew outputs.
+
+        Args:
+            all_stage_outputs (List[List[CrewOutput]]): All stage outputs.
+
+        Returns:
+            List[List[CrewOutput]]: Formatted crew outputs.
+        """
+        crew_outputs: List[CrewOutput] = [
+            output
+            for stage_outputs in all_stage_outputs[:-1]
+            for output in stage_outputs
+        ]
+        return [crew_outputs + [output] for output in all_stage_outputs[-1]]
+
+    def __rshift__(self, other: PipelineStage) -> "Pipeline":
+        """
+        Implements the >> operator to add another Stage (Crew or List[Crew]) to an existing Pipeline.
+
+        Args:
+            other (Any): The stage to add.
+
+        Returns:
+            Pipeline: A new pipeline with the added stage.
+        """
+        if isinstance(other, (Crew, Router)) or (
+            isinstance(other, list) and all(isinstance(item, Crew) for item in other)
+        ):
+            return type(self)(stages=self.stages + [other])
+        else:
+            raise TypeError(
+                f"Unsupported operand type for >>: '{type(self).__name__}' and '{type(other).__name__}'"
+            )
--- a/src/crewai/pipeline/pipeline_kickoff_result.py
+++ b/src/crewai/pipeline/pipeline_kickoff_result.py
@@ -0,0 +1,61 @@
+import json
+import uuid
+from typing import Any, Dict, List, Optional, Union
+
+from pydantic import UUID4, BaseModel, Field
+
+from crewai.crews.crew_output import CrewOutput
+from crewai.types.usage_metrics import UsageMetrics
+
+
+class PipelineKickoffResult(BaseModel):
+    """Class that represents the result of a pipeline run."""
+
+    id: UUID4 = Field(
+        default_factory=uuid.uuid4,
+        frozen=True,
+        description="Unique identifier for the object, not set by user.",
+    )
+    raw: str = Field(description="Raw output of the pipeline run", default="")
+    pydantic: Any = Field(
+        description="Pydantic output of the pipeline run", default=None
+    )
+    json_dict: Union[Dict[str, Any], None] = Field(
+        description="JSON dict output of the pipeline run", default={}
+    )
+
+    token_usage: Dict[str, UsageMetrics] = Field(
+        description="Token usage for each crew in the run"
+    )
+    trace: List[Any] = Field(
+        description="Trace of the journey of inputs through the run"
+    )
+    crews_outputs: List[CrewOutput] = Field(
+        description="Output from each crew in the run",
+        default=[],
+    )
+
+    @property
+    def json(self) -> Optional[str]:
+        if self.crews_outputs[-1].tasks_output[-1].output_format != "json":
+            raise ValueError(
+                "No JSON output found in the final task of the final crew. Please make sure to set the output_json property in the final task in your crew."
+            )
+
+        return json.dumps(self.json_dict)
+
+    def to_dict(self) -> Dict[str, Any]:
+        """Convert json_output and pydantic_output to a dictionary."""
+        output_dict = {}
+        if self.json_dict:
+            output_dict.update(self.json_dict)
+        elif self.pydantic:
+            output_dict.update(self.pydantic.model_dump())
+        return output_dict
+
+    def __str__(self):
+        if self.pydantic:
+            return str(self.pydantic)
+        if self.json_dict:
+            return str(self.json_dict)
+        return self.raw
--- a/src/crewai/pipeline/pipeline_output.py
+++ b/src/crewai/pipeline/pipeline_output.py
@@ -0,0 +1,20 @@
+import uuid
+from typing import List
+
+from pydantic import UUID4, BaseModel, Field
+
+from crewai.pipeline.pipeline_kickoff_result import PipelineKickoffResult
+
+
+class PipelineOutput(BaseModel):
+    id: UUID4 = Field(
+        default_factory=uuid.uuid4,
+        frozen=True,
+        description="Unique identifier for the object, not set by user.",
+    )
+    run_results: List[PipelineKickoffResult] = Field(
+        description="List of results for each run through the pipeline", default=[]
+    )
+
+    def add_run_result(self, result: PipelineKickoffResult):
+        self.run_results.append(result)
--- a/src/crewai/routers/init.py
+++ b/src/crewai/routers/init.py
@@ -0,0 +1 @@
+from crewai.routers.router import Router
--- a/src/crewai/routers/router.py
+++ b/src/crewai/routers/router.py
@@ -0,0 +1,60 @@
+from dataclasses import dataclass
+from typing import Any, Callable, Dict, Generic, Tuple, TypeVar
+
+from pydantic import BaseModel, Field
+
+T = TypeVar("T", bound=Dict[str, Any])
+U = TypeVar("U")
+
+
+@dataclass
+class Route(Generic[T, U]):
+    condition: Callable[[T], bool]
+    pipeline: U
+
+
+class Router(BaseModel, Generic[T, U]):
+    routes: Dict[str, Route[T, U]] = Field(
+        default_factory=dict,
+        description="Dictionary of route names to (condition, pipeline) tuples",
+    )
+    default: U = Field(..., description="Default pipeline if no conditions are met")
+
+    def __init__(self, routes: Dict[str, Route[T, U]], default: U, **data):
+        super().__init__(routes=routes, default=default, **data)
+
+    def add_route(
+        self,
+        name: str,
+        condition: Callable[[T], bool],
+        pipeline: U,
+    ) -> "Router[T, U]":
+        """
+        Add a named route with its condition and corresponding pipeline to the router.
+
+        Args:
+            name: A unique name for this route
+            condition: A function that takes a dictionary input and returns a boolean
+            pipeline: The Pipeline to execute if the condition is met
+
+        Returns:
+            The Router instance for method chaining
+        """
+        self.routes[name] = Route(condition=condition, pipeline=pipeline)
+        return self
+
+    def route(self, input_data: T) -> Tuple[U, str]:
+        """
+        Evaluate the input against the conditions and return the appropriate pipeline.
+
+        Args:
+            input_data: The input dictionary to be evaluated
+
+        Returns:
+            A tuple containing the next Pipeline to be executed and the name of the route taken
+        """
+        for name, route in self.routes.items():
+            if route.condition(input_data):
+                return route.pipeline, name
+
+        return self.default, "default"
--- a/src/crewai/task.py
+++ b/src/crewai/task.py
@@ -47,6 +47,7 @@ class Task(BaseModel):
    tools_errors: int = 0
    delegations: int = 0
    i18n: I18N = I18N()
+    name: Optional[str] = Field(default=None)
    prompt_context: Optional[str] = None
    description: str = Field(description="Description of the actual task.")
    expected_output: str = Field(
--- a/src/crewai/types/init.py
+++ b/src/crewai/types/init.py
--- a/src/crewai/types/usage_metrics.py
+++ b/src/crewai/types/usage_metrics.py
@@ -0,0 +1,36 @@
+from pydantic import BaseModel, Field
+
+
+class UsageMetrics(BaseModel):
+    """
+    Model to track usage metrics for the crew's execution.
+
+    Attributes:
+        total_tokens: Total number of tokens used.
+        prompt_tokens: Number of tokens used in prompts.
+        completion_tokens: Number of tokens used in completions.
+        successful_requests: Number of successful requests made.
+    """
+
+    total_tokens: int = Field(default=0, description="Total number of tokens used.")
+    prompt_tokens: int = Field(
+        default=0, description="Number of tokens used in prompts."
+    )
+    completion_tokens: int = Field(
+        default=0, description="Number of tokens used in completions."
+    )
+    successful_requests: int = Field(
+        default=0, description="Number of successful requests made."
+    )
+
+    def add_usage_metrics(self, usage_metrics: "UsageMetrics"):
+        """
+        Add the usage metrics from another UsageMetrics object.
+
+        Args:
+            usage_metrics (UsageMetrics): The usage metrics to add.
+        """
+        self.total_tokens += usage_metrics.total_tokens
+        self.prompt_tokens += usage_metrics.prompt_tokens
+        self.completion_tokens += usage_metrics.completion_tokens
+        self.successful_requests += usage_metrics.successful_requests
--- a/tests/crew_test.py
+++ b/tests/crew_test.py
@@ -18,6 +18,7 @@ from crewai.task import Task
 from crewai.tasks.conditional_task import ConditionalTask
 from crewai.tasks.output_format import OutputFormat
 from crewai.tasks.task_output import TaskOutput
+from crewai.types.usage_metrics import UsageMetrics
 from crewai.utilities import Logger, RPMController
 from crewai.utilities.task_output_storage_handler import TaskOutputStorageHandler

@@ -597,14 +598,10 @@ def test_crew_kickoff_usage_metrics():
    assert len(results) == len(inputs)
    for result in results:
        # Assert that all required keys are in usage_metrics and their values are not None
-        for key in [
-            "total_tokens",
-            "prompt_tokens",
-            "completion_tokens",
-            "successful_requests",
-        ]:
-            assert key in result.token_usage
-            assert result.token_usage[key] > 0
+        assert result.token_usage.total_tokens > 0
+        assert result.token_usage.prompt_tokens > 0
+        assert result.token_usage.completion_tokens > 0
+        assert result.token_usage.successful_requests > 0


 def test_agents_rpm_is_never_set_if_crew_max_RPM_is_not_set():
@@ -1315,12 +1312,12 @@ def test_agent_usage_metrics_are_captured_for_hierarchical_process():

    print(crew.usage_metrics)

-    assert crew.usage_metrics == {
-        "total_tokens": 219,
-        "prompt_tokens": 201,
-        "completion_tokens": 18,
-        "successful_requests": 1,
-    }
+    assert crew.usage_metrics == UsageMetrics(
+        total_tokens=219,
+        prompt_tokens=201,
+        completion_tokens=18,
+        successful_requests=1,
+    )


@pytest.mark.vcr(filter_headers=["authorization"])
--- a/tests/pipeline/init.py
+++ b/tests/pipeline/init.py
--- a/tests/pipeline/test_pipeline.py
+++ b/tests/pipeline/test_pipeline.py
@@ -0,0 +1,645 @@
+import json
+from unittest.mock import AsyncMock, MagicMock
+
+import pytest
+from crewai.agent import Agent
+from crewai.crew import Crew
+from crewai.crews.crew_output import CrewOutput
+from crewai.pipeline.pipeline import Pipeline
+from crewai.pipeline.pipeline_kickoff_result import PipelineKickoffResult
+from crewai.process import Process
+from crewai.routers.router import Route, Router
+from crewai.task import Task
+from crewai.tasks.task_output import TaskOutput
+from crewai.types.usage_metrics import UsageMetrics
+from pydantic import BaseModel, ValidationError
+
+DEFAULT_TOKEN_USAGE = UsageMetrics(
+    total_tokens=100, prompt_tokens=50, completion_tokens=50, successful_requests=3
+)
+
+
+@pytest.fixture
+def mock_crew_factory():
+    def _create_mock_crew(name: str, output_json_dict=None, pydantic_output=None):
+        MockCrewClass = type("MockCrew", (MagicMock, Crew), {})
+
+        class MockCrew(MockCrewClass):
+            def __deepcopy__(self, memo):
+                result = MockCrewClass()
+                result.kickoff_async = self.kickoff_async
+                result.name = self.name
+                return result
+
+        crew = MockCrew()
+        crew.name = name
+        task_output = TaskOutput(
+            description="Test task", raw="Task output", agent="Test Agent"
+        )
+        crew_output = CrewOutput(
+            raw="Test output",
+            tasks_output=[task_output],
+            token_usage=DEFAULT_TOKEN_USAGE,
+            json_dict=output_json_dict if output_json_dict else None,
+            pydantic=pydantic_output,
+        )
+
+        async def async_kickoff(inputs=None):
+            print("inputs in async_kickoff", inputs)
+            return crew_output
+
+        # Create an AsyncMock for kickoff_async
+        crew.kickoff_async = AsyncMock(side_effect=async_kickoff)
+
+        # Add more attributes that Procedure might be expecting
+        crew.verbose = False
+        crew.output_log_file = None
+        crew.max_rpm = None
+        crew.memory = False
+        crew.process = Process.sequential
+        crew.config = None
+        crew.cache = True
+
+        # Add non-empty agents and tasks
+        mock_agent = MagicMock(spec=Agent)
+        mock_task = MagicMock(spec=Task)
+        mock_task.agent = mock_agent
+        mock_task.async_execution = False
+        mock_task.context = None
+
+        crew.agents = [mock_agent]
+        crew.tasks = [mock_task]
+
+        return crew
+
+    return _create_mock_crew
+
+
+@pytest.fixture
+def mock_router_factory(mock_crew_factory):
+    def _create_mock_router():
+        crew1 = mock_crew_factory(name="Crew 1", output_json_dict={"output": "crew1"})
+        crew2 = mock_crew_factory(name="Crew 2", output_json_dict={"output": "crew2"})
+        crew3 = mock_crew_factory(name="Crew 3", output_json_dict={"output": "crew3"})
+
+        MockRouterClass = type("MockRouter", (MagicMock, Router), {})
+
+        class MockRouter(MockRouterClass):
+            def __deepcopy__(self, memo):
+                result = MockRouterClass()
+                result.route = self.route
+                return result
+            
+
+
+        mock_router = MockRouter()
+        mock_router.route = MagicMock(
+            side_effect=lambda x: (
+                (
+                    Pipeline(stages=[crew1])
+                    if x.get("score", 0) > 80
+                    else (
+                        Pipeline(stages=[crew2])
+                        if x.get("score", 0) > 50
+                        else Pipeline(stages=[crew3])
+                    )
+                ),
+                (
+                    "route1"
+                    if x.get("score", 0) > 80
+                    else "route2" if x.get("score", 0) > 50 else "default"
+                ),
+            )
+        )
+
+        return mock_router
+
+    return _create_mock_router
+
+
+def test_pipeline_initialization(mock_crew_factory):
+    """
+    Test that a Pipeline is correctly initialized with the given stages.
+    """
+    crew1 = mock_crew_factory(name="Crew 1")
+    crew2 = mock_crew_factory(name="Crew 2")
+
+    pipeline = Pipeline(stages=[crew1, crew2])
+    assert len(pipeline.stages) == 2
+    assert pipeline.stages[0] == crew1
+    assert pipeline.stages[1] == crew2
+
+
+@pytest.mark.asyncio
+async def test_pipeline_with_empty_input(mock_crew_factory):
+    """
+    Ensure the pipeline handles an empty input list correctly.
+    """
+    crew = mock_crew_factory(name="Test Crew")
+    pipeline = Pipeline(stages=[crew])
+
+    input_data = []
+    pipeline_results = await pipeline.kickoff(input_data)
+
+    assert (
+        len(pipeline_results) == 0
+    ), "Pipeline should return empty results for empty input"
+
+
+@pytest.mark.asyncio
+async def test_pipeline_process_streams_single_input(mock_crew_factory):
+    """
+    Test that Pipeline.process_streams() correctly processes a single input
+    and returns the expected CrewOutput.
+    """
+    crew_name = "Test Crew"
+    mock_crew = mock_crew_factory(name="Test Crew")
+    pipeline = Pipeline(stages=[mock_crew])
+    input_data = [{"key": "value"}]
+    pipeline_results = await pipeline.kickoff(input_data)
+
+    mock_crew.kickoff_async.assert_called_once_with(inputs={"key": "value"})
+
+    for pipeline_result in pipeline_results:
+        assert isinstance(pipeline_result, PipelineKickoffResult)
+        assert pipeline_result.raw == "Test output"
+        assert len(pipeline_result.crews_outputs) == 1
+        print("pipeline_result.token_usage", pipeline_result.token_usage)
+        assert pipeline_result.token_usage == {crew_name: DEFAULT_TOKEN_USAGE}
+        assert pipeline_result.trace == [input_data[0], "Test Crew"]
+
+
+@pytest.mark.asyncio
+async def test_pipeline_result_ordering(mock_crew_factory):
+    """
+    Ensure that results are returned in the same order as the inputs, especially with parallel processing.
+    """
+    crew1 = mock_crew_factory(name="Crew 1", output_json_dict={"output": "crew1"})
+    crew2 = mock_crew_factory(name="Crew 2", output_json_dict={"output": "crew2"})
+    crew3 = mock_crew_factory(name="Crew 3", output_json_dict={"output": "crew3"})
+
+    pipeline = Pipeline(
+        stages=[crew1, [crew2, crew3]]
+    )  # Parallel stage to test ordering
+
+    input_data = [{"id": 1}, {"id": 2}, {"id": 3}]
+    pipeline_results = await pipeline.kickoff(input_data)
+
+    assert (
+        len(pipeline_results) == 6
+    ), "Should have 2 results for each input due to the parallel final stage"
+
+    # Group results by their original input id
+    grouped_results = {}
+    for result in pipeline_results:
+        input_id = result.trace[0]["id"]
+        if input_id not in grouped_results:
+            grouped_results[input_id] = []
+        grouped_results[input_id].append(result)
+
+    # Check that we have the correct number of groups and results per group
+    assert len(grouped_results) == 3, "Should have results for each of the 3 inputs"
+    for input_id, results in grouped_results.items():
+        assert (
+            len(results) == 2
+        ), f"Each input should have 2 results, but input {input_id} has {len(results)}"
+
+    # Check the ordering and content of the results
+    for input_id in range(1, 4):
+        group = grouped_results[input_id]
+        assert group[0].trace == [
+            {"id": input_id},
+            "Crew 1",
+            "Crew 2",
+        ], f"Unexpected trace for first result of input {input_id}"
+        assert group[1].trace == [
+            {"id": input_id},
+            "Crew 1",
+            "Crew 3",
+        ], f"Unexpected trace for second result of input {input_id}"
+        assert (
+            group[0].json_dict["output"] == "crew2"
+        ), f"Unexpected output for first result of input {input_id}"
+        assert (
+            group[1].json_dict["output"] == "crew3"
+        ), f"Unexpected output for second result of input {input_id}"
+
+
+class TestPydanticOutput(BaseModel):
+    key: str
+    value: int
+
+
+@pytest.mark.asyncio
+async def test_pipeline_process_streams_single_input_pydantic_output(mock_crew_factory):
+    crew_name = "Test Crew"
+    mock_crew = mock_crew_factory(
+        name=crew_name,
+        output_json_dict=None,
+        pydantic_output=TestPydanticOutput(key="test", value=42),
+    )
+    pipeline = Pipeline(stages=[mock_crew])
+    input_data = [{"key": "value"}]
+    pipeline_results = await pipeline.kickoff(input_data)
+
+    assert len(pipeline_results) == 1
+    pipeline_result = pipeline_results[0]
+
+    print("pipeline_result.trace", pipeline_result.trace)
+
+    assert isinstance(pipeline_result, PipelineKickoffResult)
+    assert pipeline_result.raw == "Test output"
+    assert len(pipeline_result.crews_outputs) == 1
+    assert pipeline_result.token_usage == {crew_name: DEFAULT_TOKEN_USAGE}
+    print("INPUT DATA POST PROCESS", input_data)
+    assert pipeline_result.trace == [input_data[0], "Test Crew"]
+
+    assert isinstance(pipeline_result.pydantic, TestPydanticOutput)
+    assert pipeline_result.pydantic.key == "test"
+    assert pipeline_result.pydantic.value == 42
+    assert pipeline_result.json_dict is None
+
+
+@pytest.mark.asyncio
+async def test_pipeline_preserves_original_input(mock_crew_factory):
+    crew_name = "Test Crew"
+    mock_crew = mock_crew_factory(
+        name=crew_name,
+        output_json_dict={"new_key": "new_value"},
+    )
+    pipeline = Pipeline(stages=[mock_crew])
+
+    # Create a deep copy of the input data to ensure we're not comparing references
+    original_input_data = [{"key": "value", "nested": {"a": 1}}]
+    input_data = json.loads(json.dumps(original_input_data))
+
+    await pipeline.kickoff(input_data)
+
+    # Assert that the original input hasn't been modified
+    assert (
+        input_data == original_input_data
+    ), "The original input data should not be modified"
+
+    # Ensure that even nested structures haven't been modified
+    assert (
+        input_data[0]["nested"] == original_input_data[0]["nested"]
+    ), "Nested structures should not be modified"
+
+    # Verify that adding new keys to the crew output doesn't affect the original input
+    assert (
+        "new_key" not in input_data[0]
+    ), "New keys from crew output should not be added to the original input"
+
+
+@pytest.mark.asyncio
+async def test_pipeline_process_streams_multiple_inputs(mock_crew_factory):
+    """
+    Test that Pipeline.process_streams() correctly processes multiple inputs
+    and returns the expected CrewOutputs.
+    """
+    mock_crew = mock_crew_factory(name="Test Crew")
+    pipeline = Pipeline(stages=[mock_crew])
+    input_data = [{"key1": "value1"}, {"key2": "value2"}]
+    pipeline_results = await pipeline.kickoff(input_data)
+
+    assert mock_crew.kickoff_async.call_count == 2
+    assert len(pipeline_results) == 2
+    for pipeline_result in pipeline_results:
+        print("pipeline_result,", pipeline_result)
+        assert all(
+            isinstance(crew_output, CrewOutput)
+            for crew_output in pipeline_result.crews_outputs
+        )
+
+
+@pytest.mark.asyncio
+async def test_pipeline_with_parallel_stages(mock_crew_factory):
+    """
+    Test that Pipeline correctly handles parallel stages.
+    """
+    crew1 = mock_crew_factory(name="Crew 1")
+    crew2 = mock_crew_factory(name="Crew 2")
+    crew3 = mock_crew_factory(name="Crew 3")
+
+    pipeline = Pipeline(stages=[crew1, [crew2, crew3]])
+    input_data = [{"initial": "data"}]
+
+    pipeline_result = await pipeline.kickoff(input_data)
+
+    crew1.kickoff_async.assert_called_once_with(inputs={"initial": "data"})
+
+    assert len(pipeline_result) == 2
+    pipeline_result_1, pipeline_result_2 = pipeline_result
+
+    pipeline_result_1.trace = [
+        "Crew 1",
+        "Crew 2",
+    ]
+    pipeline_result_2.trace = [
+        "Crew 1",
+        "Crew 3",
+    ]
+
+    expected_token_usage = {
+        "Crew 1": DEFAULT_TOKEN_USAGE,
+        "Crew 2": DEFAULT_TOKEN_USAGE,
+        "Crew 3": DEFAULT_TOKEN_USAGE,
+    }
+
+    assert pipeline_result_1.token_usage == expected_token_usage
+    assert pipeline_result_2.token_usage == expected_token_usage
+
+
+@pytest.mark.asyncio
+async def test_pipeline_with_parallel_stages_end_in_single_stage(mock_crew_factory):
+    """
+    Test that Pipeline correctly handles parallel stages.
+    """
+    crew1 = mock_crew_factory(name="Crew 1")
+    crew2 = mock_crew_factory(name="Crew 2")
+    crew3 = mock_crew_factory(name="Crew 3")
+    crew4 = mock_crew_factory(name="Crew 4")
+
+    pipeline = Pipeline(stages=[crew1, [crew2, crew3], crew4])
+    input_data = [{"initial": "data"}]
+
+    pipeline_result = await pipeline.kickoff(input_data)
+
+    crew1.kickoff_async.assert_called_once_with(inputs={"initial": "data"})
+
+    assert len(pipeline_result) == 1
+    pipeline_result_1 = pipeline_result[0]
+
+    pipeline_result_1.trace = [
+        input_data[0],
+        "Crew 1",
+        ["Crew 2", "Crew 3"],
+        "Crew 4",
+    ]
+
+    expected_token_usage = {
+        "Crew 1": DEFAULT_TOKEN_USAGE,
+        "Crew 2": DEFAULT_TOKEN_USAGE,
+        "Crew 3": DEFAULT_TOKEN_USAGE,
+        "Crew 4": DEFAULT_TOKEN_USAGE,
+    }
+
+    assert pipeline_result_1.token_usage == expected_token_usage
+
+
+def test_pipeline_rshift_operator(mock_crew_factory):
+    """
+    Test that the >> operator correctly creates a Pipeline from Crews and lists of Crews.
+    """
+    crew1 = mock_crew_factory(name="Crew 1")
+    crew2 = mock_crew_factory(name="Crew 2")
+    crew3 = mock_crew_factory(name="Crew 3")
+
+    # Test single crew addition
+    pipeline = Pipeline(stages=[]) >> crew1
+    assert len(pipeline.stages) == 1
+    assert pipeline.stages[0] == crew1
+
+    # Test adding a list of crews
+    pipeline = Pipeline(stages=[crew1])
+    pipeline = pipeline >> [crew2, crew3]
+    print("pipeline.stages:", pipeline.stages)
+    assert len(pipeline.stages) == 2
+    assert pipeline.stages[1] == [crew2, crew3]
+
+    # Test error case: trying to shift with non-Crew object
+    with pytest.raises(TypeError):
+        pipeline >> "not a crew"
+
+
+@pytest.mark.asyncio
+async def test_pipeline_parallel_crews_to_parallel_crews(mock_crew_factory):
+    """
+    Test that feeding parallel crews to parallel crews works correctly.
+    """
+    crew1 = mock_crew_factory(name="Crew 1", output_json_dict={"output1": "crew1"})
+    crew2 = mock_crew_factory(name="Crew 2", output_json_dict={"output2": "crew2"})
+    crew3 = mock_crew_factory(name="Crew 3", output_json_dict={"output3": "crew3"})
+    crew4 = mock_crew_factory(name="Crew 4", output_json_dict={"output4": "crew4"})
+
+    pipeline = Pipeline(stages=[[crew1, crew2], [crew3, crew4]])
+
+    input_data = [{"input": "test"}]
+    pipeline_results = await pipeline.kickoff(input_data)
+
+    assert len(pipeline_results) == 2, "Should have 2 results for final parallel stage"
+
+    pipeline_result_1, pipeline_result_2 = pipeline_results
+
+    # Check the outputs
+    assert pipeline_result_1.json_dict == {"output3": "crew3"}
+    assert pipeline_result_2.json_dict == {"output4": "crew4"}
+
+    # Check the traces
+    expected_traces = [
+        [{"input": "test"}, ["Crew 1", "Crew 2"], "Crew 3"],
+        [{"input": "test"}, ["Crew 1", "Crew 2"], "Crew 4"],
+    ]
+
+    for result, expected_trace in zip(pipeline_results, expected_traces):
+        assert result.trace == expected_trace, f"Unexpected trace: {result.trace}"
+
+
+def test_pipeline_double_nesting_not_allowed(mock_crew_factory):
+    """
+    Test that double nesting in pipeline stages is not allowed.
+    """
+    crew1 = mock_crew_factory(name="Crew 1")
+    crew2 = mock_crew_factory(name="Crew 2")
+    crew3 = mock_crew_factory(name="Crew 3")
+    crew4 = mock_crew_factory(name="Crew 4")
+
+    with pytest.raises(ValidationError) as exc_info:
+        Pipeline(stages=[crew1, [[crew2, crew3], crew4]])
+
+    error_msg = str(exc_info.value)
+    print(f"Full error message: {error_msg}")  # For debugging
+    assert (
+        "Double nesting is not allowed in pipeline stages" in error_msg
+    ), f"Unexpected error message: {error_msg}"
+
+
+def test_pipeline_invalid_crew(mock_crew_factory):
+    """
+    Test that non-Crew objects are not allowed in pipeline stages.
+    """
+    crew1 = mock_crew_factory(name="Crew 1")
+    not_a_crew = "This is not a crew"
+
+    with pytest.raises(ValidationError) as exc_info:
+        Pipeline(stages=[crew1, not_a_crew])
+
+    error_msg = str(exc_info.value)
+    print(f"Full error message: {error_msg}")  # For debugging
+    assert (
+        "Expected Crew instance, Router instance, or list of Crews, got <class 'str'>"
+        in error_msg
+    ), f"Unexpected error message: {error_msg}"
+
+
+"""
+TODO: Figure out what is the proper output for a pipeline with multiple stages
+
+Options:
+- Should the final output only include the last stage's output?
+- Should the final output include the accumulation of previous stages' outputs?
+"""
+
+
+@pytest.mark.asyncio
+async def test_pipeline_data_accumulation(mock_crew_factory):
+    crew1 = mock_crew_factory(name="Crew 1", output_json_dict={"key1": "value1"})
+    crew2 = mock_crew_factory(name="Crew 2", output_json_dict={"key2": "value2"})
+
+    pipeline = Pipeline(stages=[crew1, crew2])
+    input_data = [{"initial": "data"}]
+    results = await pipeline.kickoff(input_data)
+
+    # Check that crew1 was called with only the initial input
+    crew1.kickoff_async.assert_called_once_with(inputs={"initial": "data"})
+
+    # Check that crew2 was called with the combined input from the initial data and crew1's output
+    crew2.kickoff_async.assert_called_once_with(
+        inputs={"initial": "data", "key1": "value1"}
+    )
+
+    # Check the final output
+    assert len(results) == 1
+    final_result = results[0]
+    assert final_result.json_dict == {"key2": "value2"}
+
+    # Check that the trace includes all stages
+    assert final_result.trace == [{"initial": "data"}, "Crew 1", "Crew 2"]
+
+    # Check that crews_outputs contain the correct information
+    assert len(final_result.crews_outputs) == 2
+    assert final_result.crews_outputs[0].json_dict == {"key1": "value1"}
+    assert final_result.crews_outputs[1].json_dict == {"key2": "value2"}
+
+
+@pytest.mark.asyncio
+async def test_pipeline_with_router(mock_router_factory):
+    router = mock_router_factory()
+
+    # Test high score route
+    pipeline = Pipeline(stages=[router])
+    result_high = await pipeline.kickoff([{"score": 90}])
+    assert len(result_high) == 1
+    assert result_high[0].json_dict is not None
+    assert result_high[0].json_dict["output"] == "crew1"
+    assert result_high[0].trace == [
+        {"score": 90},
+        {"route_taken": "route1"},
+        "Crew 1",
+    ]
+
+    # Test medium score route
+    pipeline = Pipeline(stages=[router])
+    result_medium = await pipeline.kickoff([{"score": 60}])
+    assert len(result_medium) == 1
+    assert result_medium[0].json_dict is not None
+    assert result_medium[0].json_dict["output"] == "crew2"
+    assert result_medium[0].trace == [
+        {"score": 60},
+        {"route_taken": "route2"},
+        "Crew 2",
+    ]
+
+    # Test low score route
+    pipeline = Pipeline(stages=[router])
+    result_low = await pipeline.kickoff([{"score": 30}])
+    assert len(result_low) == 1
+    assert result_low[0].json_dict is not None
+    assert result_low[0].json_dict["output"] == "crew3"
+    assert result_low[0].trace == [
+        {"score": 30},
+        {"route_taken": "default"},
+        "Crew 3",
+    ]
+
+
+@pytest.mark.asyncio
+async def test_router_with_multiple_inputs(mock_router_factory):
+    router = mock_router_factory()
+    pipeline = Pipeline(stages=[router])
+
+    inputs = [{"score": 90}, {"score": 60}, {"score": 30}]
+    results = await pipeline.kickoff(inputs)
+
+    print("RESULTS", results)
+
+    assert len(results) == 3
+    assert results[0].json_dict is not None
+    assert results[0].json_dict["output"] == "crew1"
+    assert results[1].json_dict is not None
+    assert results[1].json_dict["output"] == "crew2"
+    assert results[2].json_dict is not None
+    assert results[2].json_dict["output"] == "crew3"
+
+    assert results[0].trace[1]["route_taken"] == "route1"
+    assert results[1].trace[1]["route_taken"] == "route2"
+    assert results[2].trace[1]["route_taken"] == "default"
+
+
+@pytest.mark.asyncio
+async def test_pipeline_with_multiple_routers(mock_router_factory, mock_crew_factory):
+    router1 = mock_router_factory()
+    router2 = mock_router_factory()
+    final_crew = mock_crew_factory(
+        name="Final Crew", output_json_dict={"output": "final"}
+    )
+
+    pipeline = Pipeline(stages=[router1, router2, final_crew])
+
+    result = await pipeline.kickoff([{"score": 75}])
+
+    assert len(result) == 1
+    assert result[0].json_dict is not None
+    assert result[0].json_dict["output"] == "final"
+    assert (
+        len(result[0].trace) == 6
+    )  # Input, Router1, Crew2, Router2, Crew2, Final Crew
+    assert result[0].trace[1]["route_taken"] == "route2"
+    assert result[0].trace[3]["route_taken"] == "route2"
+
+
+@pytest.mark.asyncio
+async def test_router_default_route(mock_crew_factory):
+    default_crew = mock_crew_factory(
+        name="Default Crew", output_json_dict={"output": "default"}
+    )
+    router = Router(
+        routes={
+            "route1": Route(
+                condition=lambda x: False,
+                pipeline=Pipeline(stages=[mock_crew_factory(name="Never Used")]),
+            ),
+        },
+        default=Pipeline(stages=[default_crew]),
+    )
+
+    pipeline = Pipeline(stages=[router])
+    result = await pipeline.kickoff([{"score": 100}])
+
+    assert len(result) == 1
+    assert result[0].json_dict is not None
+    assert result[0].json_dict["output"] == "default"
+    assert result[0].trace[1]["route_taken"] == "default"
+
+
+@pytest.mark.asyncio
+async def test_router_with_empty_input(mock_router_factory):
+    router = mock_router_factory()
+    pipeline = Pipeline(stages=[router])
+
+    result = await pipeline.kickoff([{}])
+
+    assert len(result) == 1
+    assert result[0].json_dict is not None
+    assert result[0].json_dict["output"] == "crew3"  # Default route
+    assert result[0].trace[1]["route_taken"] == "default"
Author	SHA1	Message	Date
Brandon Hancock	c5e8302a7d	Merge branch 'feature/procedure_v2' into brandon/cre-107-pipeline-conditional-routing	2024-07-31 11:31:30 -04:00
Brandon Hancock	b51343d453	Clean up docs	2024-07-31 10:58:26 -04:00
Brandon Hancock	06350a74ef	Clean up end of docs	2024-07-31 10:57:09 -04:00
Brandon Hancock	d554e00454	Add router example to docs	2024-07-31 10:54:20 -04:00
Brandon Hancock	6aab0ebcbc	Remove overly complicated test	2024-07-31 10:12:37 -04:00
Brandon Hancock	41df368156	Merge branch 'main' into feature/procedure_v2	2024-07-31 09:49:48 -04:00
Brandon Hancock	1147a1c93e	Fix tests	2024-07-31 09:46:19 -04:00
Brandon Hancock	1154cef798	Add more tests which showed underlying issue with traces	2024-07-30 16:25:03 -04:00
Brandon Hancock	19f87f2b82	Tests working now!	2024-07-30 13:57:12 -04:00
Brandon Hancock	a79efefe7b	Fix critical circular dependency issues. Now needing to fix trace issue.	2024-07-30 13:29:01 -04:00
Brandon Hancock	4251494c55	Merge branch 'feature/procedure_v2' into brandon/cre-107-pipeline-conditional-routing	2024-07-30 09:55:20 -04:00
Brandon Hancock	072044c537	Rename variables based on joaos feedback	2024-07-30 09:54:11 -04:00
Brandon Hancock	e4b3b32b50	Merge branch 'feature/procedure_v2' into brandon/cre-107-pipeline-conditional-routing#	2024-07-30 09:44:23 -04:00
Brandon Hancock	e1a03ad97d	Change names	2024-07-30 09:43:36 -04:00
Brandon Hancock	46fdada5e4	WIP.	2024-07-30 09:36:04 -04:00
Brandon Hancock	d72f5a601e	Add missing doc string	2024-07-29 16:22:52 -04:00
Brandon Hancock	467c67c597	Merge branch 'feature/procedure_v2' into brandon/cre-107-pipeline-conditional-routing	2024-07-29 16:11:55 -04:00
Brandon Hancock	910c8df1a7	Update pipeline to use UsageMetric	2024-07-29 16:00:57 -04:00
Brandon Hancock	b9177f2d04	Merge branch 'brandon/cre-117-create-a-type-for-usage-metrics' into feature/procedure_v2	2024-07-29 15:55:11 -04:00
Brandon Hancock	03eafe1671	Fix 1 type error in pipeline	2024-07-29 15:54:49 -04:00
Brandon Hancock	f2830d9c7a	Drop todo	2024-07-29 15:51:14 -04:00
Brandon Hancock	619806f80d	Fix tests that were checking usage metrics	2024-07-29 15:48:48 -04:00
Brandon Hancock	e3182d135a	WIP. Converting usage metrics from a dict to an object	2024-07-29 15:30:54 -04:00
Brandon Hancock	3d5bbf5da9	Fix missing rag datatype	2024-07-29 15:09:59 -04:00
Brandon Hancock	75c7aaf585	Merge branch 'main' into feature/procedure_v2	2024-07-29 14:27:28 -04:00
Brandon Hancock	083949fc23	Merge branch 'main' into feature/procedure_v2	2024-07-29 12:48:53 -04:00
Brandon Hancock	04de7730fa	Add doc strings	2024-07-29 12:44:10 -04:00
Brandon Hancock	de6950046d	Add developer notes to explain what is going on in pipelines.	2024-07-29 12:36:03 -04:00
Brandon Hancock	cb2276dc7d	Add in Eduardo feedback. Still need to add in more commentary describing the design decisions for pipeline	2024-07-29 12:20:18 -04:00
Brandon Hancock	53e91a7c78	Fix circular dependencies and updated PipelineRouter	2024-07-29 11:31:34 -04:00
Brandon Hancock	cdfac165e3	WIP.	2024-07-26 17:24:09 -04:00
Brandon Hancock	31ff979a4b	In the middle of fixing router. A ton of circular dependencies. Moving over to a new design.	2024-07-26 09:24:40 -04:00
Brandon Hancock	d9e60c8b57	Drop router for now. will add in separately	2024-07-25 14:29:17 -04:00
Brandon Hancock	2119ba7c32	Starting to work on router	2024-07-25 09:33:05 -04:00
Brandon Hancock	b00bc44921	Add back in commentary at top of pipeline file	2024-07-24 11:27:47 -04:00
Brandon Hancock	6b4ebe16d0	Update Pipeline docs	2024-07-23 11:24:57 -04:00
Brandon Hancock	602ade4cc4	Update pipeline to properly process input and ouput dictionary	2024-07-23 11:12:55 -04:00
Brandon Hancock	471c5b970c	Update docs for pipeline	2024-07-22 19:28:32 -04:00
Brandon Hancock	33d9828edc	Implemented additional tests for pipeline. One test is failing. Need team support	2024-07-22 16:35:16 -04:00
Brandon Hancock	e95ef6fca9	Merge branch 'main' into feature/procedure_v2	2024-07-22 09:55:03 -04:00
Brandon Hancock	afd6bff159	Fix pipelineoutput to look more like crewoutput and taskoutput	2024-07-19 15:18:19 -04:00
Brandon Hancock	392490c48b	new pipeline flow with traces and usage metrics working. need to add more tests and make sure PipelineOutput behaves likew CrewOutput	2024-07-19 14:39:31 -04:00
Brandon Hancock	d094e178f1	Update terminology	2024-07-18 14:59:38 -04:00
Brandon Hancock	834c62feca	Going to start refactoring for pipeline_output	2024-07-18 11:20:26 -04:00
Brandon Hancock	c0c329b6e0	Merge branch 'main' into feature/procedure_v2	2024-07-17 13:37:27 -04:00
Brandon Hancock	f737b3b379	WIP	2024-07-16 11:35:24 -04:00
Brandon Hancock	467536b96a	Merge branch 'main' into feature/procedure_v2	2024-07-16 10:34:55 -04:00
Brandon Hancock	1988a00c60	Merge branch 'bugfix/langchain-tool-config-change' into feature/procedure_v2	2024-07-15 11:47:16 -04:00
Brandon Hancock	e2f4405291	Merge branch 'main' into feature/procedure_v2	2024-07-15 09:44:41 -04:00
Brandon Hancock	040e5a78d2	Add back in Gui's tool_usage fix	2024-07-15 09:21:21 -04:00
Brandon Hancock	c5002eedd9	rshift working	2024-07-12 16:48:19 -04:00
Brandon Hancock	f7680d6157	All tests are passing now	2024-07-12 11:00:29 -04:00
Brandon Hancock	adf93c91f7	WIP. Procedure appears to be working well. Working on mocking properly for tests	2024-07-11 15:26:46 -04:00