0yacinek0 / bedrock-agent

AWS bedrock tutorial

Bedrock Agent Tutorial

Refer YouTube Video: https://www.youtube.com/watch?v=6O9DqCrInvw

BytesCommerce/BytesCommerceFunction.py
This lambda function is configured in Action Group.

BytesCommerce/BytesCommerceSchema.json
This is the Open API Schema configured in Action Group.

BytesCommerce/kb-data
Here you will find two (knowledge base) files used in the demo:
ProductCatalog.csv, ProductDescriptions.pdf

BytesCommerce/InvokeAgentBytesCommerce.py
This is a lambda function to invoke the BytesCommerce Agent programmatically.
If you don't plan to invoke the Agent programmatically, you can ignore it.
You must use a version of boto3 that supports bedrock-agent-runtime (eg. version 1.34.49).

Check boto3 version in a lambda function:
print(boto3.__version__)
print(botocore.__version__)

To use latest version of boto3 with your lambda function (assuming it is not available via AWS Console by default) you can follow these steps to include it via a Layer:

1. Open Cloud Shell in AWS Console OR any shell/command-line environment with AWS configured.
   
2. Create a new directory:
   LIB_DIR=boto3-mylayer/python
mkdir -p $LIB_DIR
   
3. Install the boto3 library to LIB_DIR by running the following command:
   pip3 install boto3 -t $LIB_DIR
   
4. Zip all the dependencies to /tmp/boto3-mylayer.zip by running the following command:
   cd boto3-mylayer
zip -r /tmp/boto3-mylayer.zip .
   
5. Publish the layer by running the following command:
   aws lambda publish-layer-version --layer-name boto3-mylayer --zip-file fileb:///tmp/boto3-mylayer.zip
   The command returns the new layer's Amazon Resource Name (ARN), similar to the following one:
   arn:aws:lambda:region:$ACC_ID:layer:boto3-mylayer:1
   
6. Once the layer is created, you can go to AWS Console > lambda function (InvokeAgentBytesCommerce) and attach the layer to it.
   OR use the following command to attach the layer to your lambda function.
   aws lambda update-function-configuration --function-name <name-of-your-lambda i.e InvokeAgentBytesCommerce> --layers <layer ARN as seen above>

Within your lambda function you can print boto3 version to verify:
print(boto3.__version__)
print(botocore.__version__)

How to gain access to Foundation Model
In AWS Console > Bedrock > Foundation Models > Base Model > Model Access > Manage Model Access > Select the Models you need access to and hit Save.

Cleanup
Delete/release all resources related to Bedrock Agent - If you no longer need the Agent

• Agent (Actions and Knowledge Bases)
  
• OpenSearch Serverless Collection - This must be explicitly deleted. Although it is created automatically during Knowledge Base Creation - it doesn't get deleted automatically on Knowledge Base deletion
  
• Unsubscribe(Remove Access) to any AWS Bedrock Models (for eg. Anthropic Claude)
• Lambda function(s) & S3 bucket(s)
• Any other related resources you may have created.