SudhaarAI - Generative Code Assistant and Context-Based Code Optimizer

Let's See

SudhaarAI - Generative Code Assistant and Context-Based Code Optimizer

Writing clean, efficient, and maintainable code is a constant challenge for both students learning best practices and professionals managing complex projects. To address this, we present SudhaarAI: a context-aware, AI-powered platform that automates code analysis, generates improvements, and optimizes software projects. Unlike generic AI chatbots that lack an understanding of a developer’s specific codebase, SudhaarAI leverages a specialized Retrieval-Augmented Generation (RAG) pipeline. This system uses ChromaDB for semantic search and MongoDB for document storage, allowing the AI to deeply understand the context of the user's project by referencing authoritative programming guidelines and the user's own files. Built on a modern client-server architecture, the platform features a highly responsive frontend developed in Next.js and React. It offers developers a simple, intuitive chat interface that remembers past conversations and maintains project continuity. The backend is powered by a robust Python Flask REST API that orchestrates a multi-provider Large Language Model (LLM) ecosystem. By supporting multiple AI providers, including Groq, OpenAI, Anthropic, and local models using Ollama. SudhaarAI ensures speed, reliability, and privacy. Furthermore, the platform integrates CodeBERT to validate code quality, ensuring that recommendations are not just syntactically correct but also adhere to industry standards.

Keywords: Artificial Intelligence,Website Development,Retrieval-Augmented Generation,Large Language Models,Code Optimization,Client-Server Architecture,Software Engineering
Tools: Python,JavaScript,Nextjs,Flask,RESTAPI,Clerk Auth,MongoDB,ChromaDB
Department: Department of Computer Science
Poster
Blog
Team Members
Name Email CV
Abu Bakar abubakarp789@gmail.com
Muhammad Raqib Hayat raqibhayat263@gmail.com