Vantidge.ai is small and early. That means unusual ownership, direct client contact from the first month, and no layer of management between you and the consequences of your decisions.
What is different here
We would rather be straight about both sides than have you discover the second one in month three.
Open roles
If none of these fit but you think you should be here, write to us anyway and say why.
Own production AI systems end to end — retrieval architecture, evaluation design, deployment and the operations that keep them honest.
Build the pipelines, registries, monitoring and cost controls that turn a working model into a system someone can rely on.
Design test and evaluation harnesses for probabilistic systems, including hallucination, bias and red-team evaluation.
Own the data path in production: contracts, quality gates, lineage and the behaviour when upstream breaks.
Lead client architecture from readiness audit through production, and be the person accountable in the room.
Run pods with enough technical depth to evaluate the work, not just report on it.
How we hire
No trick questions and no unpaid project work. We are trying to find out whether you have operated a system in production, which is hard to fake and easy to demonstrate.
Step 01
Forty-five minutes with someone who will read your code, not a recruiter reading a checklist.
Step 02
A realistic use case. We ask how you would measure success and what you would refuse to ship without.
Step 03
A plausibly broken system with a boring cause. We are watching how you narrow the search space.
Step 04
Someone senior presents a plan with a real problem in it. We are explicitly scoring whether you say so.
We aim to give a decision within a week of the final conversation, and a reason either way.
Ready when you are
If you have put an AI system into production and kept it running, we want to hear from you regardless of what is listed above.