** Inspiration** Every Computer Science student faces a massive hurdle when transitioning from basic programming tutorials to real-world open-source projects on GitHub. Opening a repository with dozens of folders and thousands of lines of code is overwhelming. Beginners often don't know where the entry point is, how different files communicate, or what order to read the code in. We built RepoMentor AI to bridge this educational gap.

What it does RepoMentor AI is an intelligent web-based codebase tutor that transforms any public GitHub repository into a structured, beginner-friendly course:

Instant Architecture Overview: Summarizes the entire repository's purpose, tech stack, and workflow in plain, jargon-free language.

"Start Here" Guided Roadmap: Analyzes the file tree and recommends the exact sequence of files a beginner should read first.

Visual Flowchart Generation: Automatically generates interactive architecture and data-flow diagrams so visual learners can grasp how components connect.

Line-by-Line AI Code Tutor (Multilingual): Allows students to click on any complex file or function to get step-by-step explanations in simple English or Roman Urdu/Hindi.

Self-Assessment Quizzes: Generates quick 3-question MCQs from the repository's code to test whether the student actually understood the logic.

How we built it

Frontend: Built using React.js, HTML5, and CSS3, deployed on Vercel for a fast, responsive dark-mode developer experience.

Backend: Powered by Python and the Django REST Framework, which interacts with the GitHub REST API to fetch repository structures and file contents without cloning heavy folders.

AI Engine: Integrated with Large Language Models (Gemini API) to parse code AST/structure, generate Mermaid.js architecture diagrams, and provide contextual tutoring.