ML Platform Engineer

ML Platform Engineer / ML Infrastructure Engineer

Want to build the systems that determine how quickly frontier AI research moves?

The biggest bottleneck in modern AI isn't always the model—it's everything around it.

Training runs that fail halfway through. GPU clusters that sit underutilised. Researchers waiting on infrastructure instead of experiments.

This team is building large-scale multimodal reasoning models and agentic AI systems, with a major focus on post-training, reinforcement learning, and eventually pre-training their own frontier models. As the research organisation grows, they're investing heavily in the platform that enables researchers to move faster.

You'll join the ML Platform team responsible for the infrastructure behind every training run, evaluation pipeline, and production deployment. Your work will directly influence how quickly researchers can experiment, iterate, and ship new capabilities.

You'll work across large-scale distributed training, GPU orchestration, ML infrastructure, and inference systems—building reliable platforms that support everything from supervised fine-tuning through to large-scale GRPO and reinforcement learning workloads.

The teams scope spans:

• Building distributed training infrastructure for large language and multimodal models
• Scaling GPU clusters and improving utilisation across complex workloads
• Developing scheduling, orchestration, and fault-tolerant training systems
• Optimising inference performance across modern serving frameworks
• Building internal tooling that improves researcher productivity
• Designing reliable storage, networking, and data pipelines for ML workloads

You'll collaborate daily with Research Scientists and Research Engineers, solving the engineering problems that allow frontier models to train efficiently at scale.

You’ll be one of the engineers who has already helped scale large-scale ML workloads in production. Whether your background is ML Platform, ML Infrastructure, AI Systems, LLMOps/MLOps, distributed training, inference, or GPU infrastructure, you'll understand the engineering challenges behind training and serving frontier models—and enjoy solving them.

This is a highly technical environment where engineering quality matters as much as research. The problems are difficult, the ownership is high, and your work will have a direct impact on the pace of model development.

Package

• Location: San Francisco Bay Area or Miami hybrid
• Salary: $200,000–$300,000 base
• Bonus and meaningful stock
• Well-funded company backed by over $100M, with further funding expected
• Founded by a repeat entrepreneur with a previous billion dollar exit

If you're interested in building the infrastructure that powers the next generation of reasoning models, we'd love to hear from you.

All applicants will receive a response.