CodebaseQnA
LLM + RAG Powered Assistant

CodebaseQnA is an interactive tool I developed that is designed to streamline the onboarding process for new programmers. Instead of relying solely on static documentation, developers can directly query the codebase from their terminal to better understand structure, functionality, and implementation details.

Tools Used:

Python

Ollama

Langchain

Gradio/Rich

Features

Parses code files automatically for analysis

Generates smart embeddings based on code clusters rather than fixed length chunks

Runs locally in the CLI for fast, lightweight interaction

Supervision

Completed as a research intern under Dr. Sanaa Alwidian