Careers
We don’t post job openings
A job listing is a guess about the future — a role written before we’ve met the person who might redefine it. We’d rather work the other way around: you show us what you’ve made happen, and we design the role together.
Our hiring philosophy
Evidence over adjectives
We build AI systems whose answers can be verified — traced back to sources, checked against explicit rules. We apply the same standard to hiring.
That means we’re not interested in cover letters full of adjectives. We’re interested in things that actually happened because of you:
Something you built
A system, a repository, a tool people actually use.
Something you shipped
A product or feature that made it into the real world, not a slide deck.
Something you changed
A process you fixed, a number you moved, a customer problem you solved.
Something you led
A project, a team, a deal, a turnaround you drove to the finish line.
If your best work doesn’t fit any of these categories, send it anyway. The point isn’t the format. The point is that it’s real — and that it mattered.
Why no listings?
Most career pages list the positions a company needed six months ago. We work in a field that reinvents itself faster than job descriptions can be written.
So instead of asking “does this person fit the role?”, we ask “what becomes possible if this person joins?” The strongest people we’ve worked with never matched a listing — they made one obsolete.
If you see a gap in what we do, a capability we’re missing, or a problem you believe you can solve better than anyone else: that is the job description. Write it, and send it with your proof of work.
What we work on
We build AI for regulated industries — financial services, healthcare, legal, the public sector — where an unverifiable answer isn’t a glitch, it’s a liability. Our neurosymbolic stack makes every answer traceable and checkable, so enterprises can put AI to work where the stakes are real: decisions, audits, contracts, compliance.
This isn’t work that ends in a demo. It ends in systems running inside organizations, on their infrastructure, under their rules. If that’s the kind of impact you want your work to have, start here.
What we look for
We can't tell you the role. We can tell you the pattern
Our deep retrieval stack is built around three operations that standard RAG does not perform
The result is retrieval that behaves like a system with rules, not a model with confidence.
You finish things
Not just start them — ship them, deliver them, stand behind them.
You measure yourself by outcomes
Not effort, not activity — what changed because you were there.
You hold yourself to verifiable standards
Claims come with evidence, promises come with delivery.
You're drawn to hard constraints
Regulated industries, on-premise requirements, and European data sovereignty are not obstacles to you — they're the interesting part of the problem.
How it works
Three steps.
No forms, no portals, no keyword filters
1/
Send your proof of work
One email. Link or attach the thing you're proudest of, and tell us in a few sentences what role you think should exist for you at ExtensityAI.
