Infrastructure for AI you run yourself — observability, cost, and orchestration for local and distributed inference.
Day job: helping large customers deploy edge inference at Zenlayer. Side project: Wicklee — hardware-first observability and cost governance for self-hosted models, the tool that measures what your own GPUs actually cost. I also write about the layer above that: how systems decide where inference should run, and what a control plane for that decision would need to look like.
Based in Virginia. Currently fine-tuning a small enterprise model as a personal research project.