Summary
Alexander Wu is a member of technical staff with nine years of software engineering experience focused on large-scale ML systems and model training. Based in Berkeley, he has shipped production autoscaling and distributed scheduler work at Anyscale, scaled Ray from tens to 500+ nodes, and contributed initial Ray Data for petabyte-scale ingestion. His recent roles bridge research and engineering—working on model architecture, pretraining, and inference at Character.AI and model-training research engineering at Anthropic—reflecting a blend of systems design and generative AI performance work. He pairs deep systems instincts (OS and scheduler-level experience from CS 162 teaching) with hands-on product delivery in both startups and quantitative trading. An unconventional hobbyist streak shows in his "Gardener" entry—framing model development as GMO-free cultivation—which hints at a pragmatic, craft-focused approach to ML engineering.
9 years of coding experience