Member of Technical Staff - ML Infra

The role

You will develop algorithms and own the ML training stack for a continual learning platform. The role spans training, inference, and kernels to improve GPU scale, model size, throughput, memory efficiency, caching, and latency.

What you will do

• Build and optimize distributed training and RL infrastructure.
• Improve training experimentation throughput and shorten research iteration cycles.
• Optimize serving for production agents and training rollouts.
• Improve batching, scheduling, and KV-cache management.
• Develop and optimize GPU kernels and runtimes using CUDA, Triton, or comparable tools.
• Build reproducible benchmarks, observability, and automated research workflows.
• Work with researchers to turn new algorithms into reliable systems.

Required qualifications

• Strong fundamentals in distributed systems, networking, storage, and failure recovery.
• Experience shipping demanding systems.
• Deep specialization in one of training infrastructure, inference systems, or GPU kernels.
• Strong Python skills.
• Languages relevant to your specialty, such as C++, CUDA, or Triton.
• Understanding of PyTorch, JAX, or comparable framework internals.
• Strong profiling and debugging skills.
• Ownership and clear communication.

Nice to have

• Hands-on experience with training and RL stacks such as Miles, SkyRL, Prime Intellect’s verifiers, or comparable systems.
• Experience with vLLM, SGLang, collective communication, or ML compilers.

About the company

Trajectory is a research and product lab creating the platform for continual learning. Their research team comes from Deepmind, OpenAI, Meta Superintelligence, and product team from Figma, Apple, Stripe, and Windsurf.