Summary
Tony Wang is a Member of Technical Staff in the Bay Area with 11 years of hands-on experience at the intersection of deep learning, performance engineering, and systems design. He has moved between research and product environments—from MIT CSAIL autoscheduling for sparse tensor algebra to accelerating LLMs and production inference at companies like ASAPP, Cerebras, Neuralink, and now Anthropic. Tony combines low-level C++ and GPU/FPGA performance optimization with ML systems thinking, having even paused a PhD to try commercializing inference speedups. His GitHub projects (e.g., quokka) and a persistent developer mindset—while(true){/* passionate */}—signal continual tinkering and open-source curiosity. Comfortable in both research labs and startup scrums, he brings pragmatic engineering to hard ML systems problems. An unexpected thread: he’s worked briefly as a quantitative researcher and a salesman, giving him rare exposure to product-market and data-driven decision making.
11 years of coding experience