The part every
other agency skips.
Anyone can forward a CV. What takes work — and what we have built the software for — is finding out whether the person can actually do the job before you spend an hour with them.
-
We take the brief
A conversation with whoever will manage the hire. We want the stack, the seniority, the size and shape of the team, the things that have gone wrong in this seat before, and what would make someone fail in it. If the role is not realistic for the market or the budget, we will say so at this point rather than six weeks in.
-
We agree the rubric before we start looking
Together we settle which topics matter, how deep each should go, and what "good" looks like at this level. That rubric becomes the interview — the same one for every candidate for this role, which is the only way a comparison between two people means anything.
-
We source and do the first filter
We search, we reach out, and we read everything ourselves. Candidates who are clearly not right never reach you. Candidates who are borderline but interesting we will flag and explain rather than quietly drop.
-
Every candidate sits a structured interview
Run on CypherMeet, our own interview platform. The candidate answers out loud, is scored against the agreed rubric, and — where the role calls for it — works through a live coding exercise in a shared editor. Topic coverage is tracked, so the report can tell you what was actually examined and what was not.
-
Optional integrity checks, always with consent
If you need assurance that the person interviewing is the person applying and is working unaided, we can enable environment integrity checks. The candidate sees a clear consent screen explaining exactly what is checked before anything begins. We do not record candidates covertly, and flags go to you rather than being used to fail someone automatically.
-
You get the shortlist and the evidence
Each candidate arrives with a CV and an evaluation report: topics covered, correctness, how they handled difficulty and being stuck, the coding round if there was one, and any integrity flags. You make the decision from what they did.
What is in the report.
A score alone is useless. These are the things that let you decide.
Topic coverage
Which areas from the agreed rubric were actually examined, and which were not reached. You know the limits of the assessment, not just its verdict.
Correctness
How accurate the answers were against the rubric, with the substance of what they said — not a number floating free of any context.
Handling difficulty
What they did when they did not know something. For most senior roles this predicts performance better than whether they got every answer right.
Coding round
Where one was run: the code they wrote, an analysis of correctness and complexity, and the concerns worth raising in your own round.
Integrity flags
If checks were enabled, what was observed — reported as fact for you to weigh, never as an automatic disqualification.
Our written view
A short, plain recommendation from the person who ran the search, including our reservations. You are paying for judgement, so we give it.
Screening people is a responsibility, not just a service.
An assessment can shape someone's career. We would rather lose a client than run this badly, so these are fixed.
Consent before any monitoring
Explained in plain language, on screen, before it starts. Never covert.
The candidate sees their report
An assessment you cannot see is a gate, not feedback.
The same rubric for everyone
Applied identically for every candidate for a role — that is what makes it fair and comparable.
AI assists, humans decide
The platform scores and summarises. A person reviews every report before it leaves us.
See it on a real role.
Send us one open position. We will run the process and you can judge the output.