At Atomberg Technologies, I optimized a reverse osmosis system to increase recovery from 50% to 70%, conserving over 6,800 liters of feed water per membrane life cycle. I also analyzed 12 test configurations with Python and PFMEA to identify failure modes and an optimal setup for lifecycle testing.
On my disaster management drone capstone, I fine-tuned a YOLOv8s model for survivor detection and deployed it on Raspberry Pi 5, integrating autonomous hover-on-detection and GPS logging. At Harihar Enterprises, I designed a 6-axis grinding machine and validated its performance, helping reduce per-part cycle time to under one minute.

