Master of Software Engineering (Scalable Systems) student at Carnegie Mellon. I build inference infrastructure, backend systems, and GPU firmware recovery tools.
Project write-ups: jyeh2.github.io/projects
- 200 Bucks Inference Machine: a local LLM server built from secondhand mining GPUs. KV cache snapshots cut time to first token from 119 s to 0.3 s.
- Four Small LLMs on One 8GB Card: structured-extraction benchmark. Gemma 4 E4B reached 95% precision.
- FoodTrace: Solana farm-to-shelf provenance, HackCMU 2026 Solana track winner.
- Intel Arc B580 firmware recovery: dumping and reflashing a dead GPU's firmware.



