Text generation via Lex -> Lambda -> Bedrock

The repository shows a serverless way of how to leverage GenAI capabilities to build a NextGen chatbot.

Architecture

What resources will be created?

This CDK code will create the following:

• 1 Lex bot
• 1 Lambda (to invoke the Bedrock api and subsequently the Foundation Model provided by Anthropic to generate the text content)
• 2 Iam roles (one for the lex bot to call Lambda, one for the Lambda to call Bedrock)

Requirements

Bedrock Access

Ensure you have sufficient access to Amazon Bedrock and the different Foundation Models provided by leading AI Foundation model providers. For access, follow the documentation.

Development Environment

Cloud 9

This code for this pattern is executed in AWS Cloud9 environment. The EC2 instance used is m5.large (8 GiB RAM + 2 vCPU).

AWS setup

Region

If you have not yet run aws configure and set a default region, you must do so, or you can also run export AWS_DEFAULT_REGION=<your-region>. The region used in the demonstration is us-east-1 as Bedrock service is available only in limited regions.

Authorization

You must use a role that has sufficient permissions to create Iam roles, as well as cloudformation resources

Python >=3.8

Make sure you have python3 installed at a version >=3.8.x in the CDK environment. The demonstration has used python version 3.8 and a layer has been attached. The layer used in this demonstration has Boto3>=1.28.57 (for Bedrock service).

AWS CDK

Make sure you have the AWS CDK installed in the Cloud9 environment.

Setup

Set up environment and gather packages

cd lex-lambda-bedrock-cdk-python

Install the required dependencies (aws-cdk-lib and constructs) into your Python environment

pip install -r requirements.txt

Gather and deploy resources with the CDK

First synthesize, which executes the application, defines which resources will be created, and translates this into a cloudformation template

cdk synth

Now bootstrap, which provisions the resources you'll use when deploying the application

cdk bootstrap

and deploy with

cdk deploy

The deployment will create a Lex bot, a Lambda and a S3 bucket.

Usage

Once all the resources are created after cdk deploy finishes, go to AWS Management Console and search for Amazon Lex service. If the deployment is successful, a Lex bot named LexGenAIBot should be seen in the Bots home page.

Things to make sure in Lex console

• Click on Bot LexGenAIBot and verify that three intents Welcome Intent, GenerateTextIntent, and FallbackIntent are present as per the below screenshot.

• Click on WelcomeIntent and scroll down to Sample utterances to ensure sample utterances are created.

• Scroll down to find the Closing response section and expand the Message group dropdown. Ensure that a closing message is present.

• If all the above steps are in place, click on GenerateTextIntent available on the list of Intents (seen left-hand side).

• Once you are in the GenerateTextIntent page, scroll down to Sample utterances to ensure the utterances are created.

• Scroll down to find the Fulfilment section and click on the Advanced options button. Verify that Fulfilment Lambda code hook option is checked.

• Once the above steps are verified, go back ot the Intents page and under the Deployment section, click on Aliases.

• Click on TestBotAlias and scroll down to Languages section to find the language that used in the deployment, i.e., English (US). Click on English (US).

• The opened page shows the Lambda function for the Bot. Ensure the source Lambda and version or alias is properly set, as per the resource created from the CDK deployment.

• Once all the above steps are verified, go back to the LexGenAIBot All languages section. Click on Build and you are ready to Test the bot post successful build.

• Bot build in progress

• Once the Bot is built successfully, we're ready to test the LexGenAIBot bot. Click on the Test button

• The Test console opens up.

• Enter some sample queries and get the image generated.

Tips for best results

Keep your lambda perpetually warm by provisioning an instance for the runtime lambda

Go to Lambda console > select the function LexGenAIServerlessStack-LexGenAIBotLambda*

Versions > Publish new version

Under this version

• Provisioned Concurrency > set value to 1
• Permissions > Resource based policy statements > Add Permissions > AWS Service > Other, your-policy-name, lexv2.amazonaws.com, your-lex-bot-arn, lamdba:InvokeFunction

Go to your Lex Bot (LexGenAIBot)

Aliases > your-alias > your-language > change lambda function version or alias > change to your-version

This will keep an instance running at all times and keep your lambda ready so that you won't have cold start latency. This will cost a bit extra (https://aws.amazon.com/lambda/pricing/) so use thoughtfully.

Clean Up

To clean up the resources created as part of this demonstration, run the command cdk destroy in the directory lex-lambda-bedrock-cdk-python.