Head of Compute & Inference

# Head of Compute & InferenceRun the compute fabric that trains a 15 PB spatial foundation model and serves sub-30 ms inference at production scale.Team: High Performance ComputingLocation: San Francisco, CAType: full timeBase salary range: $350k – $600k baseApply by EmailBack to all roles## About the roleYou'll lead the compute org: from the training cluster that handles a 15 PB multi-sensor corpus, through the inference fleet serving customer queries under 30 ms, down to the dual-region failover that backs the 100% SLA on the homepage. Small team, big budget for hardware decisions that move the model architecture forward.## Responsibilities* Design and operate the training cluster for the Spatial Foundation Model.* Run the inference fleet — multi-region, autoscaled, sub-30 ms p95.* Own the cost-per-query and cost-per-training-token economics.* Recruit and grow the HPC team.## Qualifications* Director or principal-level experience running large-scale ML training and inference.* Strong networking, storage, and GPU-fabric fundamentals.* Track record of building cost-efficient inference systems.## Nice to have* Experience with sensor-data ingestion pipelines at PB scale.* Familiarity with Cloud Build / Vertex / SageMaker but not bound to any one cloud.## ApplySend a short note and your most relevant work to View email address on click.appcast.io. Salary range is $350k – $600k base for this role; recruiters calibrate within the band based on level and prior compensation.#J-18808-Ljbffr