I build full-stack products and AI systems that hold up under load, from RAG pipelines and real-time platforms to distributed backends engineered for scale.
Chemical Engineering major, Computer Applications minor, building software full-time.
Final-year engineer at NIT Tiruchirappalli who turns hard backend and AI problems into shipped products.
My work spans the stack: React and Next.js on the front, Node.js, FastAPI and Django behind it, and Redis, PostgreSQL and vector stores underneath. Lately I'm deep in applied AI: retrieval-augmented generation, hybrid search, and evaluation pipelines that keep LLMs honest.
I've shipped a cross-platform app to both app stores, built a real-time language-learning platform, and designed systems that survive high-concurrency traffic. Competitive programming keeps the fundamentals sharp, with 1100+ problems solved across LeetCode and CodeChef.
Where I've shipped: internships and product roles.
Products and systems I designed and built end-to-end.
An online language-training portal with multi-role auth, teacher discovery and course management. WebRTC + WebSockets power real-time 1:1 video sessions, messaging and file sharing for 500+ potential users, with Razorpay billing that cut manual scheduling effort by 40%.
A production-grade RAG pipeline with context-aware chunking and BGE embeddings for high-quality semantic retrieval over large corpora. A hybrid BM25 + dense search stage with cross-encoder re-ranking tunes top-k relevance, while context compression and RAGAS evaluation drive faithfulness and cut hallucinations.
A scalable ticketing system with seat-inventory management, dynamic pricing and high-concurrency support. Redis caching, distributed locking and transactional workflows prevent overselling, while optimized REST APIs deliver low-latency performance, fault tolerance and horizontal scalability.
The stack I reach for, grouped by layer.
Proof of problem-solving and continuous learning.
Across LeetCode and CodeChef, with consistent practice that keeps fundamentals and intuition sharp.
Paired with a ~1750 CodeChef rating, demonstrating strong, reliable algorithmic problem-solving.