At Agrocast Analytics Pvt Ltd, I developed a real-time drone live-streaming and AI-detection platform using YOLOv8, ByteTrack, and OSNet-ReID. It tracked 80 object classes and included auto-follow gimbal control, session logs, CSV export, and a full-stack portal.
I also deployed a multi-tenant AWS backend across two EC2 servers for client onboarding. I implemented an angle-domain pulse-and-confirm controller over the SIYI UDP SDK, using measured network loop delay and CRC16-checked packets.
At IIT Hyderabad, I ran DuckDB experiments on IMDB, TPC-H, and TPC-DS workloads and evaluated cache speedup. In my computer vision project at IIT Gandhinagar, I trained a droplet detector and used segmented tracks to estimate a 1.48 W heat-transfer rate.
My projects also include COVID data forensics and forecasting, point-cloud denoising, and an ongoing study of marketing events and stock returns. I’m pursuing a B.Tech. in Chemical Engineering with a minor in Computer Science and Engineering at IIT Gandhinagar.
