SlideFinder

Search & Build Custom Decks from Microsoft Build & Ignite Presentations

Features • Quick Start • Development

---

✨ Features

🔍 Semantic Slide Search

• Full-text search across all slide content
• AI-powered agentic retrieval via Azure AI Search
• Filter by event (Build / Ignite)
• Thumbnail previews with direct links

AI Deck Builder

• Natural language deck creation
• Multi-agent architecture (Architect, Critic, Judge)
• Automatic slide selection and curation
• Editable presentation outlines

PPTX Generation

• Merge selected slides into a single PPTX
• Preserve original formatting and styling
• Download source presentations
• Track slide origins

❤️ Favorites System

• Save slides for later
• Build collections across searches
• Persistent local storage
• Quick access sidebar

🚀 Quick Start

Prerequisites

• Python 3.11+
• Azure AI Search service
• Azure OpenAI service (for AI features)

Installation

# Clone the repository
git clone https://github.com/aymenfurter/slidefinder.git
cd slidefinder

# Install dependencies
pip install -r requirements.txt

# Configure environment
cp .env.example .env
# Edit .env with your Azure credentials

Configuration

Create a .env file with your Azure credentials:

# Azure OpenAI
AZURE_OPENAI_API_KEY=your-api-key
AZURE_OPENAI_ENDPOINT=https://your-resource.openai.azure.com/
AZURE_OPENAI_DEPLOYMENT=gpt-4.1-mini

# Azure AI Search
AZURE_SEARCH_ENDPOINT=https://your-search.search.windows.net
AZURE_SEARCH_API_KEY=your-search-key
AZURE_SEARCH_INDEX_NAME=slidefinder

Run the Application

# From project root directory
python -m uvicorn src.main:app --host 0.0.0.0 --port 7004 --reload

The app will be available at http://localhost:7004

Development

Project Structure

slidefinder/
├── src/                       # Main application code
│   ├── main.py               # FastAPI entry point
│   ├── api/
│   │   └── routes/           # API route handlers
│   │       ├── search.py     # Search endpoints
│   │       ├── slides.py     # Slide info endpoints
│   │       └── deck_builder.py
│   ├── core/                 # Configuration & utilities
│   │   ├── config.py         # Settings (pydantic-settings)
│   │   └── logging.py        # Logging setup
│   ├── models/               # Pydantic data models
│   │   ├── slide.py          # SlideInfo, SlideSearchResult
│   │   └── deck.py           # DeckSession
│   ├── services/
│   │   ├── search/           # Search service
│   │   │   └── azure.py      # Azure AI Search implementation
│   │   ├── pptx/             # PPTX generation
│   │   │   └── merger.py     # Slide merger
│   │   └── deck_builder/     # AI deck builder
│   │       ├── service.py    # Main service
│   │       ├── agents.py     # WorkflowOrchestrator
│   │       ├── workflow.py   # Multi-agent workflow
│   │       ├── prompts.py    # System prompts
│   │       ├── models.py     # Pydantic models
│   │       ├── state.py      # Workflow state
│   │       ├── helpers.py    # Utility functions
│   │       ├── events.py     # SSE event factories
│   │       ├── debug.py      # Debug event emitter
│   │       └── executors/    # Workflow executors
│   │           ├── search.py
│   │           ├── offer.py
│   │           ├── critique.py
│   │           └── judge.py
│   └── web/                  # Frontend assets
│       ├── static/           # CSS, JavaScript
│       └── templates/        # HTML templates
├── indexer/                  # Data indexing pipeline
│   ├── cli.py               # CLI for indexing
│   ├── fetcher.py           # Session data fetcher
│   ├── slide_indexer.py     # PPTX content extraction
│   ├── thumbnails.py        # Thumbnail generation
│   └── ai_search.py         # Azure AI Search upload
├── tests/                    # Test suite
├── scripts/                  # Shell scripts
├── data/                     # Runtime data (ppts, thumbnails, index)
├── infra/                    # Azure Bicep infrastructure
├── Dockerfile
├── requirements.txt
└── azure.yaml               # Azure Developer CLI config

Running Tests

# Run Python tests
python -m pytest tests/ -v

# Or use the script
./scripts/run_tests.sh

# Run JavaScript tests
cd src/web/static && npm test

Indexing Pipeline

The indexer fetches and processes slides from Microsoft Build & Ignite:

# Full pipeline
python -m indexer.cli

# Individual steps
python -m indexer.cli --step 1  # Create JSONL index
python -m indexer.cli --step 2  # Generate thumbnails
python -m indexer.cli --step 3  # Upload to Azure AI Search
python -m indexer.cli --step 4  # Verify search

Docker Deployment

# Build the image
docker build -t slidefinder-app .

# Run the container
docker run -d -p 7004:7004 --env-file .env -v $(pwd)/data:/app/data slidefinder-app

Azure Developer CLI (azd)

# Deploy to Azure
azd up

🔧 Tech Stack

Component	Technology
Backend	FastAPI, Python 3.11+, Uvicorn
Search	Azure AI Search (Agentic Retrieval)
AI	Azure OpenAI (GPT-4.1-mini)
PPTX Processing	python-pptx
Frontend	Vanilla JS, CSS3
Testing	pytest, Vitest
Deployment	Docker, Azure Container Apps

⚠️ Disclaimer

This is a personal project and is NOT affiliated with, endorsed by, or an official product of Microsoft Corporation.

All presentation content is the property of Microsoft Corporation. This tool provides search and discovery capabilities for publicly accessible conference materials.

📝 License

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