Define and instrument platform success metrics (time to first training run, job success rate, GPU utilization, cost per run).
Lead deep investigations of slow or failing jobs down to GPU profiles and hardware errors.
Partner with Infrastructure, Data Platform and Security on shared foundations.
Raise the technical bar through design reviews, proposals, code, and mentoring senior engineers.
Requirements
6+ years of professional experience with strong system design skills.
Track record building ML infrastructure end to end: training pipelines, workflow orchestration (Kubeflow, Airflow, Ray), feature pipelines, experiment tracking, model registries, model serving.
Experience designing widely adopted SDKs, APIs or developer tools.
Hands-on running GPU workloads at scale on Kubernetes: scheduling, quotas, multi-cluster dispatch, autoscaling, debugging GPU/node failures (Xid errors, NCCL hangs).
Depth in training and inference performance: distributed training, mixed precision, batching, quantization.
Experience leading cross-team technical work and clear written communication.
Bachelor's degree in Computer Science, Computer Engineering or related field, or equivalent experience.
Conditions
Office-based role, onsite Tuesday, Wednesday, Thursday with optional Monday and Friday.
Annual salary range $295,250 – $345,040 USD plus equity and benefits.