Amazon Bedrock PowerPoint Generator

An AI-powered system that automatically creates PowerPoint presentations by searching the web, generating slides, and getting the downloadable link.

Features

• Web Search: Uses Firecrawl to gather information from the internet
• PowerPoint Generation: Creates professional PPTX files with multiple slides
• Interactive Frontend: Streamlit-based chat interface
• Real-time Tracing: Shows AI agent decision-making process

Technologies Used

Backend

• Amazon Bedrock - AI agent orchestration and LLM integration
• AWS Lambda - Serverless compute for backend functions
• Amazon S3 - File storage for generated PowerPoint presentations
• Amazon API Gateway - REST API endpoints
• Python 3.9+ - Backend programming language

Frontend

• Streamlit - Interactive web interface
• AWS Amplify - Frontend hosting and deployment

Libraries & APIs

• python-pptx - PowerPoint file generation
• boto3 - AWS SDK for Python
• Firecrawl API - Web scraping and content extraction

Development & Deployment

• AWS CloudFormation - Infrastructure as Code
• Git - Version control
• Bash - Deployment scripts

Architecture

The system uses Amazon Bedrock Agents with three Lambda functions:

1. search-web - Performs web searches using Firecrawl
2. create-pptx - Generates PowerPoint files and uploads to S3

Quick Start

1. Setup AWS Resources

   • Follow instructions in setup_instructions.md
   • Deploy Lambda functions using provided code
   • Configure Bedrock agent with action groups

2. Create Lambda Layer

   ./create_lambda_layer.sh

3. Install Dependencies

   pip install -r requirements.txt

4. Run Frontend

   streamlit run chatbot_frontend.py

Usage Examples

"Research the latest AI trends and create a PowerPoint"
"Create slides about Amazon Bedrock features"
"Research quantum computing and make a presentation"

Files

• lambda_search_web.py - Web search Lambda function
• lambda_create_pptx.py - PowerPoint creation Lambda function
• chatbot_frontend.py - Streamlit web interface
• create_lambda_layer.sh - Script to create Lambda layer
• setup_instructions.md - Detailed setup guide

Requirements

• AWS Account with Bedrock access
• Firecrawl API key (set in environment variables)
• Python 3.9+
• Streamlit, boto3, python-pptx

Cost Optimization

The system uses serverless architecture to minimize costs:

• Lambda functions only run when triggered
• S3 storage for generated files
• Bedrock pay-per-use pricing

Security Notes

• Store API keys in environment variables
• Use IAM roles with minimal required permissions
• Enable MFA on AWS root account
• Consider using AWS Secrets Manager for production