Production history
Real time spent in production: networking, embedded, distributed systems, or AI that real users depend on.
We hire engineers, and not many of them
There is no standing list of openings. We add people when the work genuinely calls for it, and only people who could have scoped that work themselves.
What we look for
Engineers who have shipped, and run what they shipped. A few things they tend to have in common.
Owned the on-call
Been on the hook for what they built, and stayed to fix it when it broke.
Can hold the room
Able to sit with a client's engineering team and earn its trust, not just present to it.
Says the hard thing
Would rather flag that something is wrong than let it ship quietly.
What the work is
Real engagements, not a bench. What you would actually be doing.
On real systems
Directly on production systems: assessing AI in production, advising the team that owns it, or building inside their stack.
No bench
No waiting for assignment between engagements. You are on the work that actually needs you.
No layer in between
No management layer between you and the problem. You own the call and the outcome.
A high bar
Technical work held to a high standard, because the people doing it have built and run systems like these.
If this is the work you want
We don't keep a careers form. If you're an engineer and this is the kind of work you're looking for, write to us directly. A note about what you've built and run is worth more than a CV.