Research Engineer

As a founding Research Engineer, you'll train frontier-scale models and adapt them into specialized experts for enterprises. You will design and run experiments at scale, developing novel methods for agentic training.

You’ll work closely with researchers to experiment with and invent new algorithms, and you’ll collaborate with infrastructure engineers to post train LLMs on thousands of GPUs. We believe that research velocity is tied with having world class tooling; you’ll build tools and observability for yourself and others, enabling deeper investigation into how models specialize during training. If you get excited by challenging systems and ML problems at scale, this role is for you.

Post-train frontier scale large language models on enterprise tasks and environments

Explore cutting edge RL techniques, co-designing algorithms and systems

Partner with infrastructure engineers to scale training and inference efficiently across thousands of GPUs

Build high-performance internal tools for probing, debugging, and analyzing training runs

Experience training or serving LLMs

Experience building RL environments and evals for language models

Proficiency in PyTorch, JAX, or similar ML frameworks, and experience with distributed training

Strong experimental design skills

Background in pre- or post-training

Previous experience in high-performance computing environments or working with large-scale clusters