# 🛒 SnapCartAI > **Paste a recipe video URL → AI extracts ingredients → Adds them to your Swiggy Instamart cart** ## 🎯 What It Does 1. You paste a **YouTube recipe video URL** 2. SnapCartAI **downloads the audio** (via yt-dlp) 3. **Transcribes it** using faster-whisper (open-source Whisper) 4. **Extracts ingredients** using Groq LLM (LLaMA 3.3 70B) 5. **Searches Swiggy Instamart** via MCP and generates cart links ## 🧠 Tech Stack | Component | Technology | |-----------|-----------| | Backend | Python + FastAPI | | Audio Extraction | yt-dlp | | Transcription | faster-whisper (large-v3-turbo) | | LLM | Groq (LLaMA 3.3 70B) | | Cart | Swiggy Instamart MCP | | Frontend | HTML + CSS + Vanilla JS | ## 🚀 Quick Start ### Prerequisites - Python 3.10+ - FFmpeg (required by yt-dlp for audio extraction) - Groq API Key ([Get one free](https://console.groq.com)) ### Setup ```bash # Clone and enter project cd SnapCartAI # Install dependencies pip install -r backend/requirements.txt # Set your Groq API key # Edit backend/.env and set GROQ_API_KEY=your_key_here # Run the server python -m uvicorn backend.main:app --reload --host 0.0.0.0 --port 8000 ``` ### Open in browser ``` http://localhost:8000 ``` ## 📁 Project Structure ``` SnapCartAI/ ├── backend/ │ ├── main.py # FastAPI app │ ├── config.py # Configuration │ ├── pipeline.py # Pipeline orchestrator │ ├── services/ │ │ ├── url_parser.py # URL validation │ │ ├── audio_extractor.py # yt-dlp audio │ │ ├── transcriber.py # faster-whisper STT │ │ ├── ingredient_extractor.py # Groq LLM │ │ └── cart_service.py # Swiggy MCP │ └── requirements.txt ├── frontend/ │ ├── index.html │ ├── style.css │ └── app.js └── README.md ``` # SnapCart-AI