2
RAG based AI Teaching Assistant
A RAG-based AI Teaching Assistant built to make learning from video courses more interactive and accessible.
The system processes course videos using Whisper to generate timestamped transcripts, creates semantic embeddings using BGE-M3, and retrieves relevant sections based on a user's question. Llama 3.2 then generates an answer using the retrieved context.
A key feature is the ability to identify the relevant course video and timestamp where a topic is explained, making it easier to jump directly to the right part of the course.
Tech stack: Python, Whisper, BGE-M3, Llama 3.2, Ollama, embeddings, cosine similarity, and Retrieval-Augmented Generation (RAG).
Add a comment
0/2000