Brief
One call. The workflow, the gap, the systems the person will touch
Hire AI engineers who have shipped.
Forward-deployed and AI/ML engineers, screened by people who build these systems.
Screened by Enterprise architects with 25 years designing the systems large organizations run on.
Hiring
Who we place.
We place AI/ML engineers who have put systems into production and kept them running. Contract, contract-to-hire, or direct hire.
-
Forward-deployed engineers
Embedded with your team until the system runs on its own.
-
ML engineers
Data pipelines, evaluation, and model choices that hold up in production.
-
Agent builders
Agents that check before they act and hand the exceptions to a person.
-
AI product managers
Scope what a model decides and what a person must.
-
Domain experts
People who ran the function in finance, recruiting, or logistics.
Process
Three steps, each with a timebox.
- 01 30 minutes
- 02 Rolling
Shortlist
First profiles within two weeks, then more as they clear our screen. Two to four people who have shipped this kind of system
- 03 Your call
Embed or hire
Contract, contract-to-hire, or direct hire. Same person either way
What we screen for
Ownership of the outcome, not AI vocabulary.
Every question is about one system the candidate built and ran. We ask for specifics a resume cannot supply.
- What broke
- Can they explain what failed in production, how they found it, and what they changed?
- Who owned the result
- Did they own whether it worked for the customer, or one part of the build?
- The data underneath
- Can they describe the main table, its keys, and where the data comes from?
- How it was tested
- Can they say how many test cases there were, who labeled them, and which releases the tests blocked?
Engagement types
Three ways to bring someone on.
- Contract
- The engineer works on your team through us, for a defined phase.
- Contract-to-hire
- Start on contract. Move the engineer to your payroll when you are ready.
- Direct hire
- The engineer is your employee from day one.
All three are for roles based in the US.
Proof
Work we have done.
- Hiring Forward-Deployed Engineers 16 candidates. One cleared the bar. Five weeks screening engineers for production AI agents: what resumes hide, and the questions that reveal ownership.
- Services Human Resources Operations HR activity arrives as email. Workday needs a row. HR records that arrive as email, captured as clean rows with the source attached, then filed into Workday.
FAQ
The questions we actually get.
How do you vet engineers?
We screen for ownership of a production system, not AI vocabulary. A candidate has to explain the data under the system, how it was tested, what broke after launch, and what they did about it.
What engagement types do you offer?
Contract, contract-to-hire, and direct hire, for roles in the US.
How fast do we see a shortlist?
The brief is one 30-minute call. First profiles usually follow within two weeks. How long the search takes after that depends on the bar: in the search we published, one of sixteen candidates cleared it over five weeks.
How is this priced?
Contract: a monthly rate per person. Direct hire: a placement fee. Both discussed on the brief call.
Can a placement become a deployment, or the reverse?
Yes. Placed people work the way we do, so moving between the two is a conversation, not a restart.
Do you place outside finance, recruiting, and logistics?
Engineers and product managers, yes. Domain experts only where we've run the function ourselves.
Services
Need it built instead?
We also design and deploy AI agent systems for finance, HR, and logistics operations.
See servicesDescribe the role
Tell us who you're missing.
30 minutes. You describe the gap; we tell you whether we have the person and how fast.