SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
CuspAI is a frontier AI company focused on using artificial intelligence to discover breakthrough materials for energy, clean water, advanced computing, and carbon capture. The founding team comprises world-class researchers in AI, chemistry, and engineering, and the company is building an AI-powered 'search engine' for next-generation materials discovery.
As a Platform Engineer, you will architect and maintain the ML infrastructure platform that powers CuspAI's research teams. You'll be part of a newly created team reporting to the VP ML Engineering, with significant influence over the company's entire ML infrastructure strategy.
Key responsibilities include:
- Design and implement a cloud-native ML platform on Kubernetes that enables researchers to train, evaluate, and deploy models without infrastructure complexity
- Build and maintain the complete MLOps stack: CI/CD pipelines, model registries, experiment tracking, and deployment automation
- Architect distributed ML training infrastructure supporting multi-GPU workloads, with data pipelines, checkpointing, and resource optimization
- Ensure 99.9% uptime through monitoring, alerting, and automated recovery systems
- Manage complex multi-environment Kubernetes configurations using Kapitan
- Optimize cloud spending and resource utilization while maximizing computational throughput
- Create developer-friendly tools and abstractions enabling researchers to move from experiment to production
- Collaborate closely with ML researchers, chemists, materials scientists, and software engineers
Requirements:
- Deep expertise in building and maintaining ML infrastructure (not model development), ideally in startup environments
- Proven ability to work autonomously as a subject matter expert
- Extensive experience with ML workloads on Kubernetes and multi-cloud platforms (AWS, GCP)
- Strong Python and/or Go programming skills for infrastructure automation
- Expert-level proficiency in Infrastructure as Code (Terraform, Helm, Kapitan, GitOps)
- Experience operating distributed ML platforms for JAX and/or PyTorch
- GPU infrastructure experience, specifically managing GPU clusters and performance optimization
- Previous background as a Software Engineer before moving into DevOps/MLOps, or regular coding practice
- Personal interest in AI applications for materials science and sustainability
- Passion for enabling scientists to work on world-changing challenges
Bonus qualifications:
- Experience with Flyte, Kapitan, or Pants
- HPC environment and job scheduler experience (Slurm, PBS)
- ML serving infrastructure knowledge (Triton, TorchServe, KServe)
- Scientific computing workflows and data management familiarity
- Multi-tenant ML platform experience
- Background supporting research teams
