Hardware and firmware engineer on an NIH BRAIN Initiative–funded project at Prof. Shadi Dayeh's IEBL lab, UC San Diego. Built a portable, wireless 6,144-channel neural acquisition system for epilepsy patients — comprising a custom carrier card, SerDes depth links, and surface implant adapter, all orchestrated by a Xilinx K26 SoM running PetaLinux and OpenEphys. Validated in animal trials.
View Project →
Designed a 180nm CMOS operational amplifier capable of over 80 dB of gain and 35 MHz of bandwidth for UCSD ECE 164 (Fall 2022). The architecture combines a folded cascode with a common source amplifier stage, biased via a constant-transconductance reference circuit with current mirroring. Design verified in Cadence Virtuoso. Selected to present from ~50 participants and awarded 2nd place by an Apple engineers panel.
View Project →
Research into using Unity3D as a versatile synthetic data generation platform for Advanced Driver Assistance Systems. Built an automated pipeline leveraging Unity's scripting and physics engine to produce RGB-D images, Velodyne-like point clouds, 2D/3D bounding boxes, and pixel-accurate ground truth — all without expensive real-world data collection. A single RTX 2080 Ti achieves 5 Hz generation with full environmental control over lighting, weather, and scene context. Preliminary results show synthetic data boosts 2D object detection performance with the same generalizing capability as real datasets.
View Project →