An AI-powered learning assistant that lets you chat with YouTube videos, extract transcripts, and listen to transcripts. Inspired by Google NotebookLM, built with FastAPI, LangChain, Gemini Pro, Streamlit & Docker.
- Extract YouTube video transcripts
- Ask contextual questions using Gemini Pro LLM (LangChain)
- Convert transcripts to audio with gTTS
- Interactive Streamlit frontend
- FastAPI backend handling all AI logic
- Fully Dockerized & deployment ready
Please add .env file to backend folder content - GEMINI_API_KEY = "YOUR API KEY"
git clone https://github.com/AbuZar-Ansarii/NotebookLLM-YouTube-Video-AI-Assistant.git
cd youtube-learning-assistant
# Docker pull command
docker pull abuzar718/aivideoassistant-backend:latest
docker pull abuzar718/aivideoassistant-frontend:latest
## 📂 Project Structure
youtube-learning-assistant/
├── backend/ # FastAPI API (LLM logic here)
├── frontend/ # Streamlit UI (API calls only)
├── Dockerfile.backend
├── Dockerfile.frontend
├── docker-compose.yml
├── requirements.txt
├── .env.example
└── README.md
---