I'm Eklakh Dewan, an AI Engineer and Data Science undergraduate at KPR Institute of Engineering and Technology (KPRIET), Coimbatore. I build systems that go past the notebook — full pipelines that take raw data in, and return a served, explainable prediction out the other end.

My work centers on NLP, agentic LLM workflows, and explainable machine learning — recommendation engines that show their reasoning with SHAP, multilingual chatbots with real intent pipelines, and FastAPI services built for actual inference load, not demos.

During a Data Science internship at Flowrage Technology, I shipped ML applications end-to-end, optimized inference latency by roughly 20%, and improved model accuracy by roughly 12% through feature engineering and tuning — while keeping the codebase modular enough for the next person to extend it.

Outside coursework, I'm comfortable working alongside AI coding tools (Copilot, Claude, Cursor) to move fast under deadline pressure — recently representing KPRIET at HackVega 2.0, a national-level individual hackathon spanning aptitude, technical coding, and live AI-assisted coding rounds.