DevPulse

home page
repositary
profile
dashboard
project health

Inspiration

Modern software projects are becoming increasingly complex, making it difficult for developers to quickly understand unfamiliar codebases, debug issues, and maintain up-to-date documentation. We wanted to build an AI-powered assistant that acts like a software engineering teammate—helping developers explore repositories, explain code, generate documentation, and improve overall project quality. Our goal was to combine intelligent local code analysis with OpenAI models while minimizing token usage and API costs.

What it does

DevPulse AI is an AI-powered Software Engineering Command Center that enables developers to upload a project repository and interact with it using natural language. It provides AI-powered code explanations, automated documentation generation, bug analysis, project health insights, and repository search. Instead of manually reading thousands of lines of code, developers can simply ask questions and receive context-aware answers.

How we built it

We built the frontend using React and Vite, creating a clean and intuitive user interface. The backend was developed with Python FastAPI, which handles repository uploads, local code parsing, Retrieval-Augmented Generation (RAG), and communication with the OpenAI API. We used ChromaDB to store embeddings for semantic search and SQLite for lightweight data storage. The application first performs local analysis to reduce unnecessary API calls, then retrieves only the most relevant code snippets before sending them to OpenAI, significantly reducing token usage while improving response quality.

During development, we used OpenAI Codex to accelerate code generation, scaffold backend modules, and improve development speed. We also used GPT-5.6 to refine prompts, optimize the application architecture, improve documentation, and enhance the overall user experience.

Challenges we ran into

One of our biggest challenges was designing an efficient backend that minimizes OpenAI token consumption while still delivering accurate responses. We also worked on creating a modular architecture, implementing RAG for code retrieval, and integrating multiple components such as repository parsing, embeddings, and AI services into a seamless workflow.

Accomplishments that we're proud of

Built a complete AI-powered software engineering assistant.
Successfully integrated OpenAI models with Retrieval-Augmented Generation (RAG).
Designed a token-efficient architecture using local analysis and semantic search.
Developed automated documentation generation and intelligent repository analysis.
Created a modern, responsive interface with a scalable backend architecture.

What we learned

This project helped us gain hands-on experience with FastAPI, React, Retrieval-Augmented Generation (RAG), vector databases, prompt engineering, and OpenAI APIs. We also learned the importance of designing AI systems that are both efficient and cost-effective by reducing unnecessary token usage through local processing and intelligent retrieval.

What's next for DevPulse AI

Our future roadmap includes GitHub integration for automatic repository synchronization, team collaboration features, pull request reviews, CI/CD integration, real-time project monitoring, support for multiple programming languages, cloud deployment, and advanced AI-powered code review capabilities. Our vision is to make DevPulse AI an intelligent software engineering companion that assists developers throughout the entire software development lifecycle.