Build and productionize real-time ML systems for consumer AI: serve and optimize large-scale diffusion and LLM inference, improve model performance (quantization, distillation, pruning), implement personalization, ranking, and moderation, scale GPU infrastructure, run experiments and A/B tests, and own reliability, observability, and cost-performance tradeoffs.

Own and operate the end-to-end creation platform (frontend, backend, infra, models). Improve real-time generation latency and reliability, optimize cost at scale, maintain queues/fallbacks, and ship continuous improvements to the core create/remix loop and feed/ranking systems.