Staff ML Infrastructure Engineer

Staff / Lead ML Infrastructure Engineer

San Francisco, CA — Onsite

Salary - Over market average + equity

We are building one of the world's leading generative video and multimodal AI platforms, and we're looking for a senior infrastructure engineer to drive the backbone that makes it possible. This role is ideal for an engineer from a top-tier tech company who has built cloud-scale systems, high-performance compute platforms, and battle-tested CI/CD pipelines that support complex ML workloads.

What You'll Own

• Core ML Platform Architecture: Design and evolve the infrastructure that supports large-scale generative video and multimodal model training, evaluation, and deployment.
• High-Throughput Compute Systems: Build and optimize GPU/TPU clusters, distributed training systems, and orchestration layers tailored for video-heavy pipelines.
• Production Reliability for Generative Models: Create the tooling and services needed to safely push frequent model updates while handling massive compute loads and long-running jobs.
• End-to-End CI/CD for ML: Lead the development of automated pipelines for model training, validation, artifact management, and production rollout.
• Multimodal Data Infrastructure: Build systems to ingest, version, transform, and serve large-scale video, audio, and text datasets with high reliability.
• Internal Developer Experience: Partner with research, product, and applied ML teams to build intuitive internal tooling for experiment tracking, model lineage, and resource scheduling.
• Technical Leadership: Mentor engineers, set platform standards, and influence long-term architectural direction.

What You've Done

• Experience architecting and operating large-scale infrastructure at a cloud provider, hyperscaler, or leading AI company.
• Built or owned mission-critical CI/CD systems, high-capacity compute platforms, or data infrastructure supporting ML teams.
• Deep experience with distributed compute across GPUs/accelerators, Kubernetes, and cloud infrastructure (AWS/GCP/Azure).
• Strong engineering fundamentals in Python, Go, or equivalent languages.
• Previous exposure to ML training pipelines—especially systems that handle heavy video, multimodal, or high-dimensional data.
• Demonstrated ability to lead complex cross-org initiatives and drive technical strategy.

Nice to Have

• Experience with video processing systems, large-scale media pipelines, or streaming architectures.
• Familiarity with modern multimodal or video-generation frameworks (PyTorch, JAX, diffusers, custom accelerators).
• Experience with Ray, Triton, CUDA optimization, or specialized scheduling for ML workloads.
• Background working in high-growth AI startups or research-focused environments.
• Security and compliance considerations for models that generate or process user content.

Why Join

• Shape the underlying platform powering one of the most advanced generative video systems in the world.
• Influence the future of multimodal AI by building infrastructure that directly accelerates research and product breakthroughs.
• Work closely with experienced founding engineers, researchers, and platform builders from leading tech companies.
• Highly competitive compensation, meaningful equity, and strong in-person engineering culture in San Francisco.