AWS PowerPoint Script Generator

An intelligent presentation script generator that analyzes PowerPoint slides and creates natural, professional presentation scripts using AWS Bedrock and Claude 3.7 Sonnet with advanced caching and real-time AWS Documentation MCP integration.

🎯 Problem Statement

AWS Solutions Architects face a recurring challenge when preparing for customer presentations, technical sessions, and internal briefings. The process of creating engaging, accurate, and well-timed presentation scripts is both time-consuming and complex, requiring:

Current Pain Points

Time-Intensive Process: Manual script writing takes 2-4 hours per presentation
Technical Accuracy Concerns: Ensuring AWS service information is current and accurate
Inconsistent Quality: Script quality varies based on individual SA experience and available time
Language Barriers: Need for localized scripts for global audiences (Korean, English, etc.)
Time Management Issues: Difficulty in allocating appropriate time per slide based on content complexity
Repetitive Work: Similar presentations require recreating scripts from scratch

Business Impact

Reduced Productivity: SAs spend valuable time on script preparation instead of customer engagement
Inconsistent Customer Experience: Varying presentation quality across different SAs
Missed Opportunities: Rushed script preparation leads to suboptimal customer presentations
Scalability Challenges: Manual process doesn't scale with growing SA team demands

💡 Solution Overview

The AWS PowerPoint Script Generator addresses these challenges through an intelligent, automated approach that leverages cutting-edge AI technologies to transform PowerPoint presentations into professional, ready-to-deliver scripts.

Key Capabilities

Multimodal AI Analysis: Analyzes both text content and visual elements in slides
Intelligent Time Allocation: Dynamically assigns presentation time based on slide complexity and importance
Real-time AWS Documentation: Integrates with AWS Documentation MCP server for technical accuracy
Multi-Language Support: Generates scripts in Korean and English with natural localization
Persona-Aware Generation: Adapts to individual SA presentation styles and confidence levels
Real-Time Processing: Generates comprehensive scripts in minutes, not hours

Business Value

Time Savings: Reduces script preparation time from 2-4 hours to 5-10 minutes (75-90% reduction)
Quality Consistency: Maintains professional presentation standards across all SAs
Technical Accuracy: Real-time AWS documentation integration ensures current information
Global Scalability: Multi-language support enables consistent quality worldwide
Cost Efficiency: Frees up SA time for high-value customer engagement activities
Knowledge Democratization: Junior SAs can produce senior-level presentation quality

✨ Key Features

🔍 Intelligent Slide Analysis

Multimodal Processing: Analyzes both text content and visual elements
AWS Service Detection: Automatically identifies and categorizes AWS services mentioned
Content Summarization: Extracts key concepts and themes from each slide
Technical Complexity Assessment: Evaluates presentation difficulty level

🧠 AI-Powered Script Generation

Claude 3.7 Sonnet Integration: Advanced language model for natural script creation
Prompt Caching: Optimized performance with intelligent caching strategies
Persona-Aware Generation: Adapts to presenter's style and confidence level
Slide Flow Continuity: Ensures smooth transitions between slides

📚 Real-time AWS Knowledge Enhancement

MCP Integration: Real-time AWS documentation retrieval via Model Context Protocol
Service-Specific Information: Detailed AWS service descriptions and best practices
Fallback Knowledge Base: Comprehensive offline AWS service database for reliability
Technical Accuracy: Ensures accurate AWS terminology and concepts with live documentation

🎨 Customizable Output

Multiple Script Styles: Technical, conversational, or formal presentation styles
Timing Guidance: Precise timing recommendations for each slide
Speaker Notes: Detailed presentation tips and guidance
Q&A Preparation: Anticipated questions and suggested answers

🏗️ Architecture

Core Components

┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐
│   Streamlit     │    │   Multimodal     │    │   Script        │
│   Frontend      │───▶│   Analyzer       │───▶│   Generator     │
└─────────────────┘    └──────────────────┘    └─────────────────┘
                                │                        │
                                ▼                        ▼
                       ┌──────────────────┐    ┌─────────────────┐
                       │   Real MCP       │    │   Cache         │
                       │   Client         │    │   Manager       │
                       │   (AWS Docs)     │    │   (Multi-layer) │
                       └──────────────────┘    └─────────────────┘

Technology Stack

Frontend: Streamlit with responsive UI components
AI Models: Claude 3.7 Sonnet via AWS Bedrock
Document Processing: python-pptx, Pillow for image processing
MCP Integration: Real-time AWS Documentation MCP server client
Caching: Multi-layer caching with TTL management
Logging: Structured logging with loguru

🚀 Quick Start

Prerequisites

Python 3.10 or higher (required for AWS Documentation MCP server compatibility)
AWS CLI configured with appropriate permissions
Internet connection for AWS Bedrock and MCP services

⚡ One-Command Installation (Recommended)

The fastest and most reliable way to get started:

# Clone the repository
git clone https://github.com/jesamkim/aws-pptx-script-generator.git
cd aws-pptx-script-generator

# Complete automated installation with environment validation
./manage.sh install

# Start the application
./manage.sh start

That's it! The application will be available at http://localhost:8501

🛠️ Management Commands

The manage.sh script provides comprehensive project management:

# Installation and Setup
./manage.sh install     # Full installation with advanced environment setup
./manage.sh validate    # Validate environment configuration
./manage.sh repair      # Automatically repair environment issues

# Application Management
./manage.sh start       # Start the Streamlit application
./manage.sh stop        # Stop the running application
./manage.sh restart     # Restart the application
./manage.sh status      # Show application and environment status

# Maintenance and Testing
./manage.sh test        # Run comprehensive test suite
./manage.sh update      # Update dependencies
./manage.sh clean       # Clean temporary files and caches
./manage.sh logs        # Show application logs

# Backup and Recovery
./manage.sh backup      # Create backup of configuration
./manage.sh restore     # Restore configuration from backup
./manage.sh uninstall   # Remove virtual environment and configs

# Help
./manage.sh help        # Show all available commands

🔧 Advanced Installation Options

Platform-Specific Scripts

# Linux/macOS - Basic installation
./install.sh

# Windows - PowerShell installation
.\install.ps1

# Advanced environment setup (Linux/macOS)
./scripts/setup_environment.sh

Manual Installation

Click to expand manual installation steps

Clone Repository

git clone https://github.com/jesamkim/aws-pptx-script-generator.git
cd aws-pptx-script-generator

Set up Python Environment

python -m venv aws-venv
source aws-venv/bin/activate  # Linux/macOS
# aws-venv\Scripts\activate   # Windows

Install Dependencies

pip install -r requirements.txt
pip install awslabs-aws-documentation-mcp-server pydantic-settings

Configure AWS Credentials

aws configure
# Enter: Access Key, Secret Key, Region (us-west-2), Format (json)

Start Application
```
streamlit run streamlit_app.py
```

📖 Usage Guide

Step-by-Step Process

1. Upload PowerPoint File

Upload your .pptx file using the drag-and-drop interface or file uploader.

2. AI Analysis Results

The system performs multimodal analysis of the uploaded PowerPoint and displays the analysis results including slide content, AWS services detected, and technical complexity assessment.

3. Presenter Information

Enter your presenter information including name, title, and other personal details.

4. Presentation Settings

Configure presentation parameters including language selection, presentation duration, target audience, and other settings.

5. Generated Script Results

View the final generated presentation script with dynamic timing allocation and comprehensive content.

Generation Options

Basic Cached Generation

Fast script generation with prompt caching
Suitable for standard presentations
Optimized for performance

Optimized Agent Generation

Advanced multi-agent workflow
Enhanced AWS knowledge integration
Parallel processing for complex presentations
Real-time MCP documentation retrieval

🧪 Testing and Validation

Automated Testing Suite

# Run all tests
./manage.sh test

# Individual test categories
python tests/test_installation.py      # Installation verification
python tests/test_mcp_connection.py    # MCP integration test
python tests/test_mcp_integration.py   # Comprehensive MCP test
python tests/test_script_generation.py # Script generation test
python tests/run_tests.py             # Complete test runner

Environment Validation

# Comprehensive environment check
./manage.sh validate

# Advanced validation with detailed report
python scripts/validate_environment.py

Test Categories

Installation Tests: Verify all components are properly installed
MCP Integration Tests: Test real-time AWS documentation retrieval
Unit Tests: Individual component testing
Integration Tests: End-to-end workflow validation
Performance Tests: Caching and optimization validation

🔧 Configuration

AWS Permissions

Required AWS IAM permissions:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "bedrock:InvokeModel",
        "bedrock:InvokeModelWithResponseStream"
      ],
      "Resource": [
        "arn:aws:bedrock:*::foundation-model/us.anthropic.claude-3-7-sonnet-*"
      ]
    }
  ]
}

Bedrock Model Access

Access AWS Console → Navigate to AWS Bedrock service
Request Model Access → Go to "Model access" and request access to "Anthropic Claude 3.7 Sonnet"
Wait for Approval → Usually approved within 24 hours

Environment Configuration

The application uses these configuration files:

.env: Environment variables (auto-generated during installation)
mcp-settings.json: MCP server configuration (auto-generated)
config/aws_config.py: AWS Bedrock client configuration
config/mcp_config.py: MCP client configuration

🔍 Troubleshooting

Automated Problem Resolution

# Automatic diagnosis and repair
./manage.sh validate    # Identify issues
./manage.sh repair      # Automatically fix common problems
./manage.sh status      # Check current status

Common Issues and Solutions

Environment Issues

# Python version problems
./manage.sh validate    # Check Python version
# Follow installation guide for Python 3.10+

# Missing dependencies
./manage.sh repair      # Auto-install missing packages

# Virtual environment issues
./manage.sh install     # Recreate environment

AWS Configuration Issues

# AWS credentials not configured
aws configure

# Bedrock access issues
aws bedrock list-foundation-models --region us-west-2

# Region configuration
export AWS_REGION=us-west-2

MCP Integration Issues

# Test MCP connection
python tests/test_mcp_connection.py

# Verify MCP server installation
pip show awslabs-aws-documentation-mcp-server

# Check MCP configuration
cat mcp-settings.json

Advanced Troubleshooting Tools

# Detailed environment validation
python scripts/validate_environment.py

# View comprehensive logs
./manage.sh logs

# Clean and reset environment
./manage.sh clean
./manage.sh install

📊 Performance Optimization

Caching Strategy

The application implements multi-layer caching:

Prompt Caching: Claude API prompt caching for repeated requests
Response Caching: In-memory caching of AI responses
MCP Caching: AWS documentation caching with TTL
Slide Analysis Caching: Cached multimodal analysis results

Performance Metrics

Script Generation: 20-60 seconds depending on complexity
Cache Hit Rate: 60-80% for repeated presentations
MCP Response Time: 3-8 seconds per AWS service query
Memory Usage: ~200-500MB during processing

🤝 Development

Development Setup

Clone and set up environment
Install development dependencies
```
pip install -r requirements-dev.txt
```
Run tests to verify setup
```
./manage.sh test
```

Code Quality Standards

Follow PEP 8 guidelines
Use type hints for all functions
Add docstrings for public methods
Maintain comprehensive test coverage

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🆕 What's New in v2.1

✨ Advanced Environment Management

Comprehensive Management Tool: ./manage.sh with 15+ commands for complete lifecycle management
Automated Environment Setup: scripts/setup_environment.sh with OS detection and automatic dependency installation
Intelligent Validation: scripts/validate_environment.py with 8-point comprehensive environment checking
Automatic Repair: scripts/repair_environment.py for automatic issue resolution

🧪 Enhanced Testing and Validation

Comprehensive Test Suite: Complete test coverage with automated test runner
Real-time MCP Testing: Live AWS Documentation MCP server connectivity testing
Environment Validation: Automated verification of all system components
Performance Monitoring: Built-in performance metrics and optimization tracking

🔧 Improved User Experience

One-Command Installation: Complete setup with ./manage.sh install
Intelligent Problem Resolution: Automatic diagnosis and repair of common issues
Comprehensive Status Monitoring: Real-time application and environment status
Backup and Recovery: Configuration backup and restore capabilities

🚀 Production-Ready Features

Advanced Error Handling: Comprehensive error detection and recovery
Logging and Monitoring: Structured logging with multiple log levels
Configuration Management: Automated configuration file generation and validation
Performance Optimization: Multi-layer caching and optimization strategies

Version: 2.1.0
Last Updated: July 31, 2025
Compatibility: Python 3.10+, AWS Bedrock, Claude 3.7 Sonnet, AWS Documentation MCP Server

🔗 MCP Integration Status: ✅ ACTIVE - Real-time AWS documentation retrieval enabled
🛠️ Management Tools: ✅ ACTIVE - Complete automated environment management
🧪 Testing Suite: ✅ ACTIVE - Comprehensive validation and testing framework

Name		Name	Last commit message	Last commit date
Latest commit History 15 Commits
config		config
docs		docs
images		images
scripts		scripts
src		src
tests		tests
.gitignore		.gitignore
LICENSE		LICENSE
QUICKSTART.md		QUICKSTART.md
README.md		README.md
design.md		design.md
install.ps1		install.ps1
install.sh		install.sh
manage.sh		manage.sh
mcp-settings.json		mcp-settings.json
requirements-dev.txt		requirements-dev.txt
requirements.md		requirements.md
requirements.txt		requirements.txt
streamlit_app.py		streamlit_app.py

License

jesamkim/aws-pptx-script-generator

Folders and files

Latest commit

History

Repository files navigation