Senior Software Engineer
CurrentMachine Learning Training Platform: • Working on Pinterest's Next-Gen ML dedicated Kubernetes cluster foundation. • Empowered 300+ ML Engineers scale up their ML models in training and inference. • Built orchestration layer on top of open-source Kubeflow to host 5000+ daily ML jobs. • Owner of Ray infrastructure at Pinterest, unlocked many advanced and efficient ML use cases. • Led Jupyter team (3 FTE), accountable for 1000+ Monthly Active Users internal users. • Managed JupyterHub on K8s to allow engineers to prototype ideas at ease. • Developed tools to speed up ML Engineers' model development cycle. • Designed ML integration tests to ensure confidence and safety in ML changes. • Delivered multiple company-wide PyTorch version major upgrades for training and serving.