00 — AI & Technology Recruitment
Hiring for AI-shaped roles is hard because the job descriptions are new and the candidates who've genuinely done them are rare. We recruit for roles we work in ourselves — that's the difference.
01 — Two front doors
02 — The roles
Build
People who have taken models from notebook to production and back again — retrieval pipelines, fine-tuning, evaluation harnesses — and know which parts of the hype cycle to ignore.
Run
GPU scheduling, inference serving, observability, security for systems that answer with your company’s most sensitive documents. The roles that separate a demo from an asset.
Feed it
Engineers who get data out of the systems it’s trapped in and keep it honest — pipelines, governance, quality checks — because a private AI is only as good as what it can read.
Also: AI-savvy product managers, solutions architects who talk to business owners, and the operations people who keep both honest. If a role sounds unusual, that's often exactly why it needs a recruiter who understands it.
03 — From our point of view
“You can't staff a revolution with a job board. The people who've built private AI know what 'has actually done it' looks like.”
04 — Next step
Either way it starts with one honest conversation about what the job really is — or what you're really looking for.
Talk to Dircel →