AI/ML Engineer, Research Intern at IIT Madras, and Open Source Contributor passionate about building intelligent systems, AI agents, and developer-focused tools.

Currently pursuing an M.Tech in Data Science & AI at CUSAT and actively contributing to open-source projects while building real-world AI applications.

• 🔨 Building: OSSCopilot — an AI-powered open-source contribution assistant
• 🤖 Interests: AI Agents, RAG Systems, LLM Applications, Applied ML
• 📚 Research: Transfer Learning, Fault Diagnosis, Deep Learning
• 🌐 Backend: FastAPI, PostgreSQL, APIs, Docker
• 🚀 Open Source: PatchPilot, ArchNode, ILoveAgents

"Because finding the right open-source issue to contribute to shouldn't be a full-time job."

An AI-powered assistant that matches contributors to GitHub issues based on their skill profile. Uses semantic embeddings to surface the most relevant issues — not just keyword matches.

• GitHub OAuth, skill extraction via Claude API, semantic issue matching via OpenAI embeddings
• pgvector on Neon for vector search, Next.js 15 frontend in a Linear/Supabase dark aesthetic
• Stack: Next.js 15 · FastAPI · PostgreSQL + pgvector · Claude API · OpenAI Embeddings · Neon

A RAG-based document Q&A system. Upload docs, ask questions, get grounded answers — not hallucinations.

• LangChain / LlamaIndex pipeline, Chroma vector store, OpenAI / Anthropic APIs
• Stack: Python · LangChain · Chroma · FastAPI

• PatchPilot — Built REST APIs and contributed a machine-learning severity ranking pipeline for scan prioritization.
• ArchNode — Implemented centralized date normalization and ingestion reliability improvements.
• ILoveAgents — Added timeout handling for LLM provider requests to improve reliability of agent workflows.

Languages

AI / ML

Backend

Frontend

Tools

Open to collabs, hackathons, and PRs that actually pass CI.