Setup Amazon Bedrock agent, knowledge base, and action group with Streamlit

Table of Contents

1. Introduction
2. Prerequisites
3. Diagram
4. Configuration and Setup
   • Step 1: Creating S3 Buckets
   • Step 2: Knowledge Base Setup in Bedrock Agent
   • Step 3: Lambda Function Configuration
   • Step 4: Setup Bedrock Agent and Action Group
5. Step 5: Setup Knowledge Base with Bedrock Agent
6. Step 6: Create an Alias
7. Step 7: Testing the Setup
   • Testing the Knowledge Base
   • Testing the Bedrock Agent
8. Step 8: Setup and Run Streamlit App on EC2 (Optional)
9. Cleanup
10. Security
11. License

Overview

In this project, we will set up an Amazon Bedrock agent with an action group that dynamically creates an investment company portfolio based on specific parameters. The agent also has Q&A capabilities for Federal Open Market Committee (FOMC) reports, leveraging a Streamlit framework for the user interface. Additionally, this exercise includes a method for sending emails, although the email functionality will not be fully configured.

For those who prefer an Infrastructure-as-Code (IaC) solution, we provide an AWS CloudFormation template that will deploy most of the necessary resources, including S3 buckets, an action group, and a Lambda function. You will still need to manually create the knowledge base with the already provided resources, but the steps are relatively straightforward. If you would like to deploy this workflow via AWS CloudFormation, please refer to the workshop guide here.

Alternatively, this README will walk you through the step-by-step process to set up the Amazon Bedrock agent manually using the AWS Console.

Prerequisites

• An active AWS Account.
• Familiarity with AWS services like Amazon Bedrock, S3, Lambda, and Cloud9.

Diagram

Configuration and Setup

Step 1: Creating S3 Buckets

• Please make sure that you are in the us-west-2 region. If another region is required, you will need to update the region variable theRegion in the invoke_agent.py file code.
• Domain Data Bucket: Create an S3 bucket to store the domain data. For example, call the S3 bucket knowledgebase-bedrock-agent-{alias}. We will use the default settings.

• Next, we will download the domain data from here. On your computer, open terminal or command prompt, and run the following curl commands to download the data:

• For Mac

    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20230201.pdf --output ~/Documents/fomcminutes20230201.pdf
  
    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20230322.pdf --output ~/Documents/fomcminutes20230322.pdf
    
    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20230614.pdf --output ~/Documents/fomcminutes20230614.pdf
    
    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20230726.pdf --output ~/Documents/fomcminutes20230726.pdf
    
    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20230920.pdf --output ~/Documents/fomcminutes20230920.pdf
    
    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20231101.pdf --output ~/Documents/fomcminutes20231101.pdf
  

• For Windows

    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20230201.pdf --output %USERPROFILE%\Documents\fomcminutes20230201.pdf
    
    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20230322.pdf --output %USERPROFILE%\Documents\fomcminutes20230322.pdf
    
    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20230614.pdf --output %USERPROFILE%\Documents\fomcminutes20230614.pdf
    
    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20230726.pdf --output %USERPROFILE%\Documents\fomcminutes20230726.pdf
    
    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20230920.pdf --output %USERPROFILE%\Documents\fomcminutes20230920.pdf
    
    curl https://raw.githubusercontent.com/build-on-aws/bedrock-agents-streamlit/main/S3docs/fomcminutes20231101.pdf --output %USERPROFILE%\Documents\fomcminutes20231101.pdf

• Also, you have the option to download the .pdf files from here. These files will download to your Documents folder. Upload these files to S3 bucket knowledgebase-bedrock-agent-{alias}. These files are the Federal Open Market Committee documents describing monetary policy decisions made at the Federal Reserved board meetings. The documents include discussions of economic conditions, policy directives to the Federal Reserve Bank of New York for open market operations, and votes on the federal funds rate. More information can be found here. Once uploaded, please select one of the documents to open and review the content.

Step 2: Knowledge Base Setup in Bedrock Agent

• Before we setup the knowledge base, we will need to grant access to the models that will be needed for our Bedrock agent. Navigate to the Amazon Bedrock console, then on the left of the screen, scroll down and select Model access. On the right, select the orange Manage model access button.

• Select the checkbox for the base model columns Amazon: Titan Embeddings G1 - Text and Anthropic: Claude 3 Haiku. This will provide you access to the required models. After, scroll down to the bottom right and select Request model access.

• After, verify that the Access status of the Models are green with Access granted.

• Now, we will create a knowledge base by selecting Knowledge base on the left, then selecting the orange button Create knowledge base.

• You can use the default name, or enter in your own. Then, select Next at the bottom right of the screen.

• Sync S3 bucket knowledgebase-bedrock-agent-{alias} to this knowledge base.

• For the embedding model, choose Amazon: Titan Embeddings G1 - Text. Leave the other options as default, and scroll down to select Next.

• On the next screen, review your work, then select Create knowledge base (Creating the knowledge base may take a few minutes. Please wait for it to finish before going to the next step.)

• When the knowledge base is complete, you will see a green message at the top similar to the following:

Step 3: Lambda Function Configuration

• Create a Lambda function (Python 3.12) for the Bedrock agent's action group. We will call this Lambda function PortfolioCreator-actions.

• Copy the python code provided below, or from the file here into your Lambda function.

import json

def lambda_handler(event, context):
    print(event)
  
    # Mock data for demonstration purposes
    company_data = [
        #Technology Industry
        {"companyId": 1, "companyName": "TechStashNova Inc.", "industrySector": "Technology", "revenue": 10000, "expenses": 3000, "profit": 7000, "employees": 10},
        {"companyId": 2, "companyName": "QuantumPirateLeap Technologies", "industrySector": "Technology", "revenue": 20000, "expenses": 4000, "profit": 16000, "employees": 10},
        {"companyId": 3, "companyName": "CyberCipherSecure IT", "industrySector": "Technology", "revenue": 30000, "expenses": 5000, "profit": 25000, "employees": 10},
        {"companyId": 4, "companyName": "DigitalMyricalDreams Gaming", "industrySector": "Technology", "revenue": 40000, "expenses": 6000, "profit": 34000, "employees": 10},
        {"companyId": 5, "companyName": "NanoMedNoLand Pharmaceuticals", "industrySector": "Technology", "revenue": 50000, "expenses": 7000, "profit": 43000, "employees": 10},
        {"companyId": 6, "companyName": "RoboSuperBombTech Industries", "industrySector": "Technology", "revenue": 60000, "expenses": 8000, "profit": 52000, "employees": 12},
        {"companyId": 7, "companyName": "FuturePastNet Solutions", "industrySector": "Technology",  "revenue": 60000, "expenses": 9000, "profit": 51000, "employees": 10},
        {"companyId": 8, "companyName": "InnovativeCreativeAI Corp", "industrySector": "Technology", "revenue": 65000