🎥 YouTube Notes RAG System

A Retrieval-Augmented Generation (RAG) system that lets you:

• Fetch YouTube video transcripts
• Generate embeddings and store them locally
• Query videos using natural language
• Get AI-powered summaries and Q&A

Features

• Transcript Extraction: Automatically fetch YouTube video transcripts
• Local Vector Database: ChromaDB for efficient similarity search
• LLM Integration: Ollama with local LLMs (DeepSeek, Llama3, etc.)
• Strict Context-Only Answers: Prevents hallucinations
• Modular Architecture: Easily swap components (database, LLM, etc.)

Tech Stack

• Backend: Python 3.10+
• Vector DB: ChromaDB
• Embeddings: SentenceTransformers (all-MiniLM-L6-v2)
• LLM: Ollama (local models)
• UI: Streamlit

Installation

1. Prerequisites:

   • Ollama installed and running
   • Python 3.10+

2. Set up virtual environment:

   python -m venv .venv
   source .venv/bin/activate  # Linux/Mac
   # .venv\Scripts\activate  # Windows
   

3. Install dependencies:

    pip install -r requirements.txt
   

4. Download LLM model:

   ollama pull deepseek-r1:1.5b
   

Project Structure

youtube-notes/
├── database/           
│   ├── databaseInterface.py    
│   └── chroma.py       
├── utils/
│   ├── youtube_utils.py 
├── youtube_notes.py    
├── app.py              
└── requirements.txt

Usage

1. Run the app:

    streamlit run app.py
   

2. Process a video:

   Enter YouTube Video ID (e.g., pNJUyol15Jw)

   Click "Process Video" to generate embeddings and summary

3. Ask questions:

   Type natural language questions about the video

   Get answers strictly based on the transcript