Adding Bedrock provider example for memory embedder

[peterableda]

Adding example on how to use Amazon Bedrock models as an embedder for memory.

[peterableda]

If there is a better/easier way, let me know, I dug the chromadb and crewai codebase for an hour to get this working.

[joaomdmoura]

Disclaimer: This review was made by a crew of AI Agents.

Code Review Report for PR #1933

Overview

The changes add documentation for using Amazon Bedrock embeddings with crewAI's memory system. The patch modifies docs/concepts/memory.mdx by adding a new section with configuration examples.

Documentation Quality Analysis

Positive Aspects

1. The example is well-structured and includes essential components.
2. An important prerequisite note about boto3 installation is included.
3. Clear demonstration of environment variable usage.
4. Complete configuration structure for Bedrock embeddings.

Issues and Recommendations

1. Indentation Consistency

The code example has inconsistent indentation in the embedder configuration.

Current:

    embedder={
	    "provider": "bedrock",
		"config":{
			"session": boto3_session,
			"model": "amazon.titan-embed-text-v2:0",
			"vector_dimension": 1024
		}
	}

Recommended Fix:

    embedder={
        "provider": "bedrock",
        "config": {
            "session": boto3_session,
            "model": "amazon.titan-embed-text-v2:0",
            "vector_dimension": 1024
        }
    }

2. Missing Error Handling Documentation

Consider adding a note about potential errors and how to handle them:

Recommended Addition:

"""
Note: Ensure all environment variables are properly set before initializing the boto3 session.
The code may raise boto3.exceptions.NoCredentialsError if credentials are missing.

Error handling example:
try:
    boto3_session = boto3.Session(...)
except boto3.exceptions.NoCredentialsError:
    print("AWS credentials not found. Please check your environment variables.")
"""

3. Environment Variables Documentation

Add a section explaining required environment variables:

Required Environment Variables:

- AWS_REGION_NAME: Your AWS region (e.g., 'us-east-1')
- AWS_ACCESS_KEY_ID: Your AWS access key
- AWS_SECRET_ACCESS_KEY: Your AWS secret key

4. Code Style Improvements

The ellipsis usage for agents and tasks could be more informative:

Current:

agents=[...],
tasks=[...],

Recommended:

agents=[
    # Add your agents here
    # Example: Agent(name="researcher", ...)
],
tasks=[
    # Add your tasks here
    # Example: Task(description="research task", ...)
],

5. Missing Type Hints

Consider adding type hints for better code documentation:

Recommended Addition:

from typing import Optional
from boto3.session import Session

boto3_session: Session = boto3.Session(
    region_name: str = os.environ.get("AWS_REGION_NAME"),
    aws_access_key_id: str = os.environ.get("AWS_ACCESS_KEY_ID"),
    aws_secret_access_key: str = os.environ.get("AWS_SECRET_ACCESS_KEY")
)

Security Considerations

1. Add a warning about secure credential management.
2. Recommend using AWS credential management best practices.
3. Suggest using AWS IAM roles when possible.

Additional Recommendations

1. Add a section about supported Bedrock models.
2. Include information about pricing and usage limits.
3. Provide a complete working example with actual agents and tasks.
4. Add a troubleshooting section for common issues.

Overall Assessment

The documentation addition is valuable and well-structured but could benefit from more comprehensive error handling, security considerations, and practical examples. The current implementation is functional but could be enhanced with the suggested improvements for better user experience and maintainability.

Suggested Final Changes:

1. Fix indentation in the code example.
2. Add security and error handling sections.
3. Expand environment variables documentation.
4. Include type hints.
5. Add practical usage examples.

[bhancockio]

Thank you for adding this!