Koppile Sannam is a full-stack AI project designed to bring an intelligent — and occasionally spicy — twist to the world of code reviews. It analyzes GitHub repositories using advanced AI models, offering precise feedback that’s brutally honest yet technically accurate.
Created by Alfie Varghese, this platform merges machine learning with a distinct South Indian flavor of humor through its virtual persona — Professor Pisharody. The professor delivers insights in a unique blend of wit, sarcasm, and engineering wisdom, ensuring developers get both a laugh and a lesson.
Every roast is powered by a backend built with Node.js and Express, integrated with the GitHub API and Gemini AI for deep repository analysis. The frontend, hosted on Vercel, uses React with a custom “Leaked Document” theme for a polished and immersive experience.
What makes Koppile Sannam stand out is its dual nature — a technical tool that genuinely helps developers improve, and a creative platform that turns those improvements into a story worth sharing.
At its core, the goal is simple: to make learning from your code mistakes fun, relatable, and unforgettable. Whether your repository is elegant or an “ethum potti”, the Professor will find a way to make it a teaching moment.
The term “Koppile Sannam” loosely translates to “what’s cooking inside the head” — a playful expression of curiosity and mischief. It represents the project’s mission: to peek into developers’ code and serve feedback that’s equal parts wisdom and wit.