Lead productionization and optimization of ML models for a global, serverless inference platform. Build benchmarking/evaluation tooling, improve inference performance (quantization, compilation, batching), integrate models with heterogeneous GPUs and accelerators, enhance deployment workflows and observability, and mentor engineers while driving ML production best practices.

Design and build a global, serverless AI inference platform running on heterogeneous GPUs and accelerators. Optimize model scheduling, request routing, caching, and load balancing for low latency. Improve reliability and observability, lead cross-functional projects, and mentor junior engineers.