AI · RAG
Oct 2025
FilingIQ
Ask an 800-page IPO filing a question in plain English. It answers, with citations.
Recall@5 0.85
MRR 0.62
cited 83%
Final-year B.Tech EEE at NIT Trichy, CGPA 8.48. I build backend systems and applied AI — RAG pipelines, LLM orchestration, and APIs that hold up under real traffic.
Every number on this page came off a real eval set or a real trace.
Market-1501 · zero-shot.
Ask an 800-page IPO filing a question in plain English. It answers, with citations.
A multi-turn conversational AI backend built to survive real traffic — cached, rate-limited, streamed.
A voice assistant that runs entirely on your machine. Nothing leaves the device.
Four agents read your CloudWatch logs and argue until they find the bug, then drop the RCA into Slack.