Overview

Workshop Overview

In this workshop, you will build a complete AI Chatbot using serverless architecture on AWS. The chatbot uses Claude Haiku 4.5 deployed entirely serverless with low costs and automatic scaling capabilities.

Modules

Module 1: Setup Amazon Bedrock Module 2: Create Lambda Function Module 3: Configure API Gateway Module 4: Deploy Frontend to S3 Module 5: Testing & Debugging

Architecture Overview

alt text

AWS Services & Their Roles in the Workshop

1. Amazon Bedrock — AI Brain

  • Process and generate responses from user messages.
  • Uses Claude Haiku 4.5 (Messages API).
  • Features:
    • Inference Profiles
    • Conversation history management
    • Temperature control

2. AWS Lambda — Backend Logic

Role:

  • Receive requests from API Gateway
  • Validate input, format prompt
  • Call Bedrock and return response

Features:

  • Node.js 24 runtime
  • Environment variables (MODEL_ID, ALLOWED_ORIGIN)
  • CloudWatch Logs
  • Input validation + error handling
  • Retry logic
  • CORS handling

3. Amazon API Gateway — API Endpoint

Role:

  • Public REST API for browsers
  • CORS management
  • Route POST /chat to Lambda
  • Throttling to prevent abuse

Features:

  • REST API or HTTP API
  • CORS configuration
  • Lambda Proxy Integration
  • OPTIONS method for CORS preflight

4. Amazon S3 — Frontend Hosting

  • Static Website Hosting
  • Host HTML/CSS/JS
  • Allow users to access chatbot via browser

5. Amazon CloudWatch — Monitoring & Debugging

  • Collect logs from Lambda
  • Debug request/response & errors

Workshop Learning Objectives

After the workshop, you will:

  • Understand serverless architecture and how services interact
  • Deploy a production-ready chatbot with Claude Haiku 4.5
  • Integrate Bedrock API with Lambda
  • Configure API Gateway
  • Host static website on S3
  • Debug using CloudWatch Logs
  • Optimize serverless costs