1 day ago
Berlin, Germany +2 moreSenior
Responsibilities
- Design and implement a cloud-native Kubernetes platform for training, evaluating, and deploying ML models.
- Own the MLOps stack, including CI/CD pipelines, model registries, experiment tracking, and deployment automation.
- Build infrastructure for distributed ML training across GPU clusters, including data pipelines, checkpointing, and resource optimization.
- Maintain 99.9% platform uptime through monitoring, alerting, and automated recovery systems.
- Manage multi-environment Kubernetes configurations across development, staging, and production using Kapitan.
- Optimize cloud resource usage and spending while maximizing computational throughput.
- Create developer tools and abstractions that help researchers move experiments into production.
- Collaborate with ML researchers, chemists, materials scientists, and software engineers to meet infrastructure needs.
Requirements
- Deep experience building and maintaining ML infrastructure rather than developing models, ideally in startup environments.
- Deep experience with ML workloads on Kubernetes and multi-cloud platforms including AWS, GCP, and neoclouds.
- Strong Python and/or Go programming skills for infrastructure automation.
- Expert-level Infrastructure as Code experience with Terraform, Helm, Kapitan, and GitOps workflows.
- Experience operating distributed ML platforms for JAX and/or PyTorch; TensorFlow experience is less relevant.
- Experience managing GPU infrastructure, GPU clusters, and performance optimization.
- Previous software engineering experience or regular coding experience in professional or personal projects.
- Ability to work autonomously as a subject matter expert and provide best-practice guidance.
- Interest in enabling scientific work in materials science, energy, clean water, compute, and carbon capture.
- Preferred experience with Flyte, Kapitan, Pants, or similar tools; HPC environments and Slurm or PBS; ML serving infrastructure such as Triton, TorchServe, or KServe; scientific computing workflows; multi-tenant ML platforms; or research-team support.
Benefits
- Competitive salary and equity in CuspAI.
- 28 days of holiday in Germany, the Netherlands, and the UK, or 21 days in Japan, Singapore, and the US, in addition to local public holidays.
- 26 weeks of fully paid primary-caregiver parental leave and 12 weeks of fully paid secondary-caregiver parental leave.
- Professional development budget.
- Interdisciplinary collaboration with AI researchers, computational chemists, experimental scientists, and engineers.
- Equal-opportunity workplace with reasonable interview and workplace adjustments available.
Tech Stack
Categories
About CuspAI
CuspAI builds an AI-driven materials discovery platform that helps R&D teams in chemicals, energy, and manufacturing design and evaluate novel molecules and materials. The company offers software and models accessed via cloud APIs and collaborations to prioritize candidates, guide experiments, and shorten lab cycles. Founded in 2024 and headquartered in Cambridge, it is a privately held startup focused on applying generative and physics-informed AI to sustainability and next‑generation materials problems.