Discipline
Compute and GPU infrastructure recruiters.
The engineers who design, deploy, and run the GPU clusters behind modern AI, from bring-up to scale.
How this market works
What this search really turns on.
Scale is not a resume line
Running a handful of nodes and running thousands of GPUs are different jobs. We screen on the size and topology a candidate has actually operated, because that is what predicts whether they can do yours.
The stack is specific
NVIDIA reference architectures, InfiniBand, NCCL, the scheduler, the whole toolchain: knowing the acronyms is not the same as having debugged them at 3am under a deployment deadline. We can tell the difference.
Bring-up and steady-state are different skills
Standing a cluster up is not the same as keeping it healthy for a year. We are clear with both sides about which problem a candidate has really solved.
The talent pool is tiny and busy
The people who have done this at scale are rare and are almost never on the market. Reaching them takes direct, credible outreach, not a job post.
Plain answers
Questions, answered straight.
What roles does this cover?
Infrastructure and platform engineers, cluster and HPC engineers, and the architects and leads who own compute at scale.
Do you place site reliability and operations too?
Yes. The people who keep large clusters observable and online are part of this practice, and we recruit for them alongside the build-side roles.
Standing up compute, or ready to run it at scale?
Tell us what you are building. The first conversation costs nothing and stays between us.