YouTube RAG Q&A with Google Gemini

🎬 YouTube Video Q&A powered by Retrieval-Augmented Generation (RAG) and Google Gemini LLM.
Ask questions about YouTube videos and get answers directly from the transcript.

---

Introduction

The YouTube RAG Q&A with Google Gemini project is an interactive web application that allows users to ask questions about YouTube videos and get precise answers based solely on the video transcript.

Using a Retrieval-Augmented Generation (RAG) approach, the app splits video transcripts into smaller chunks, converts them into embeddings, and stores them in a vector database (FAISS). When a user asks a question, the system retrieves the most relevant chunks and uses Google Gemini LLM to generate context-aware, accurate responses.

This project demonstrates the combination of video processing, natural language understanding, and large language models to create an intelligent and interactive video assistant.

---

Live Demo

🌐 Try the app

---

📸 Screenshot

---

Features

• Automatic video validation and transcript fetching from YouTube.
• RAG-based retrieval: answers are generated using only the video transcript.
• Powered by Google Gemini LLM for high-quality answers.
• Supports multiple languages in transcripts (e.g., English & Hindi).
• Built with Streamlit for an interactive web interface.

---

How It Works

1. Video Validation: The app checks if the YouTube URL is valid and accessible.
2. Transcript Fetching: Captions are fetched using YouTubeTranscriptApi.
3. Text Chunking: The transcript is split into smaller chunks for better retrieval.
4. Embeddings & Vector Store: Each chunk is converted into embeddings using Google Gemini, and stored in FAISS.
5. Retrieval-Augmented Generation (RAG): When you ask a question, relevant chunks are retrieved from FAISS.
6. LLM Answering: Google Gemini LLM generates answers based only on the retrieved transcript chunks.

---

How to Use

1. Enter your Google AI Studio API key in the input field.
   • To get a free API key: Google AI Studio.
2. Paste the full YouTube video URL you want to analyze.
3. Type your question about the video.
4. Click Submit Question to get the AI-generated answer.

---

Installation

1. Clone the repository:

git clone https://github.com/tahirkorma/rag-youtube-assistant.git
cd <rag-youtube-assistant>

2. Create a virtual environment (optional but recommended):

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

3. Install dependencies:

pip install -r requirements.txt

4. Run the App

streamlit run app.py

Requirements

Python 3.8+
Streamlit
Pytube
youtube-transcript-api
LangChain
FAISS
langchain-google-genai
(All dependencies are listed in requirements.txt)