One bar, held across the network.
The standard doesn't relax for an internal hire or tighten for an external one. The same bar applies whether the engineer ends up on an Orion team, a partner company, or a recruiter's pool.
Shortlist is a verification engine: a senior-reviewed bar for what good means in a domain, and the regional pool that bar produces. It defines the standard, holds engineers to it, and connects the verified pool to the work that needs it.
AI now produces code, completes take-homes, and clones writing style. Resumes are noisier than they have been in twenty years. Phone screens never caught identity in the first place. Without a standard, senior is a self-reported word.
The engine exists because the signal under engineering hiring broke and nothing replaced it. A senior in the field, on the record, looking at real work with the candidate in the room — that's the missing rung. The standard is the way it goes back in.
The standard doesn't relax for an internal hire or tighten for an external one. The same bar applies whether the engineer ends up on an Orion team, a partner company, or a recruiter's pool.
Every verified profile carries the same kind of receipt — assessment score, lab artifact, SME endorsement on record, identity confirmation. Profiles read side by side without disclaimers.
What good means in cloud infrastructure is not what good means in applied ML. The bar is set by a senior in each domain, not borrowed from a generic interview rubric. A domain is added when a senior in it can run the standard.
AI calibrates the bar, scores at scale, and surfaces signal that's harder to see by hand. A senior in the field signs off on whether the engineer cleared it. No model decides alone.
Clearing the standard takes about two and a half hours of real work: a domain assessment, a take-home lab, and a live session with a senior engineer in the field. Identity is re-verified at placement, so the engineer verified is the engineer who starts.
The live session carries the weight — a lab shows what someone can produce; a senior in the room shows whether they understand it.
The standard is the product. Everything else runs on top of it.
A senior isn't checking whether an answer can be produced — tools produce answers. They're checking whether the engineer can reason about their own work: why this design, where it broke, what they'd change.
AI assists where signal is harder to see by hand: calibrating the bar against a domain, scoring at scale, surfacing patterns across attempts. A senior signs off on whether the engineer cleared the bar. No model decides alone.
The standard proves a real engineer did real work and could defend it to someone senior. It doesn't claim more than that.
A senior-reviewed bar per domain. What good means in cloud infrastructure differs from what good means in applied ML. The standard is calibrated by seniors in the field — not borrowed from a generic rubric.
A system of signal: assessment tooling, hands-on labs, live sessions with senior engineers in the field, identity checks. AI assists in calibration and scoring; a human signs off. No model decides alone.
Every verified engineer carries a profile — domain scores, lab artifacts, SME endorsement, identity confirmation. Built from results, not from self-reported claims.
The profile connects to where the work is — Orion's own teams, partner companies, recruiter rails. The engine routes the match; the human keeps the relationship.
When a match converts, identity is re-confirmed. The engineer verified is the engineer who starts.
End-to-end across candidate onboarding, evaluation, partner matching, and placement — all in the admin console.
Recorded in the actual admin console. End-to-end candidate and partner onboarding UAT-passed; AI matching live.
The engine routes verified engineers to the work that matches; the human relationship stays with the partner or recruiter.
The teams building Orion's own platforms draw from the pool.
Organizations Orion works with draw verified engineers through a relationship, not a listings page.
Recruiters and talent teams run on the verification as infrastructure — the engine supplies the proof; the recruiter keeps the relationship.
Senior through lead and engineering management — the standard is about depth a title can't show, so there are no junior tiers. A domain is added when a senior in it can run the standard.
The aim isn't throughput; it's the floor. Starting in Northern Virginia, the engine is building a verified pool deep enough to shift what incoming tech talent means in the region — a quality bar that compounds as it grows. The pool is built before the demand: verify the engineers, hold the standard, and let the work find the proof.
Live roles from employers using the engine.
It does one thing: make verified mean something, and connect what it means to the work that needs it.