← All roles

Member of Technical Staff - ML Infra

Trajectory· Member of Technical Staff
Apply on company site
San FranciscoFull-timePosted 2w ago

About the role

ML Infrastructure Engineer specializing in the ML training stack, spanning training, inference, and kernels. Focus on improving the continual learning platform’s training stack with strong ownership and cross-stack impact.

Responsibilities

  • Build and optimize distributed training and RL infrastructure, including rollout execution, GPU scheduling, training parallelisms, checkpointing, and recovery.
  • Improve training experimentation throughput and shorten research iteration cycles for 100B+ parameter models.
  • Optimize serving for production agents and training rollouts, including batching, scheduling, and KV-cache management.
  • Balance latency, throughput, cost, and model quality in inference.
  • Develop and optimize GPU kernels and runtimes using CUDA, Triton, or comparable tools; improve memory use and execution efficiency while preserving numerical correctness.
  • Build reproducible benchmarks, observability, and automated research workflows that propose changes, run experiments, and validate improvements.
  • Collaborate with researchers to turn new algorithms into reliable systems across training, inference, and kernels.

Qualifications

  • Strong fundamentals in distributed systems, networking, storage, and failure recovery, with experience shipping demanding systems.
  • Required deep specialization in one of training infrastructure, inference systems, or GPU kernels, supported by relevant systems or optimizations.
  • Strong Python skills and languages relevant to specialty (e.g., C++, CUDA, or Triton).
  • Understanding of PyTorch, JAX, or comparable framework internals, with strong profiling and debugging skills.
  • Ownership and clear communication: work closely with researchers and deliver measurable performance gains while preserving correctness and reliability.

Skills

Required:Distributed systemsSpecialization in training infra / inference / GPUsPythonC++ / CUDA / TritonFramework internals
Last updated Oct 14, 2026. Interested? Apply on the company site.

About Trajectory

The Platform for Continual Learning

Research lab and product company building the platform for continual learning.

Continual LearningCustomersField NotesPlatformSecurityTeamVisionbessemer-ai-infrastructure-2026
Company profileWebsite

More open roles at Trajectory