Senior Engineer, ML Infrastructure · Tesla AI · Palo Alto
I build data platforms, schedulers, and cluster-efficiency systems for Tesla AI. My work supports data operations for FSD, Optimus, and Digital Optimus across large GPU fleets.
Tesla AI · internal platform
Own a large-scale MapReduce system used for data acquisition, curation, mining, auto-labeling, training, and evaluation. It runs across 10 GPU clusters, processing 10B+ jobs per month and parallelizing an average of 3,000 years of compute.
Tesla AI · GPU infrastructure
Built placement and live-defragmentation systems that compact allocations into contiguous topology while clusters remain online. The scheduler operates across a heterogeneous GPU fleet.
Tesla AI · FSD · Optimus · Digital Optimus
Build pipelines that turn large-scale data acquisition into training-ready datasets, with lineage across artifacts, datasets, training jobs, and models.
Built an edge-to-cloud data broker processing 10B+ manufacturing data points per day across Austin, Berlin, and Fremont. Developed a 5M writes/second durable queue and a distributed cache that cut storage by 80% across 50 deployments.
Researched formal verification for autonomous systems under DARPA ARCOS with Lockheed Martin. Co-authored PSY-TaLiRo and published eight peer-reviewed papers.
Built the image-ranking model at the core of AfterShoot's photo-culling product, now used by approximately 250,000 photographers.