We're Building round1

Hiring's first round should produce evidence, not opinions. And you shouldn't pay for it until you hire the right person.

Those two sentences are what round1 is being built on. Here is what that means in practice: round1 runs the first round for you, and hands back a report of scores against the rubric you defined, where every score quotes the moment in the interview it came from. You are charged when someone starts, and not before.

The rest of this is why we think it needs to exist.

Two options, both of which end in a guess

Today you have two ways to fill a role.

You can use an agency. They charge 20 to 30% of first-year salary, they take weeks, and what arrives is a shortlist with a recommendation attached. You are trusting somebody else's read of a conversation you were not in.

Or you post the role and cast a wide net. Now you have volume, and the screening is yours. Somebody on your team spends 30 to 45 minutes per candidate on calls, learning the things a CV was never going to tell them. Those calls happen when your team has time, which is why good candidates go cold waiting.

Both roads end in the same place. A decision made on evidence that is thin and subjective, filtered through somebody's impression of somebody else before it reaches the person who actually has to choose.

An interview that isn't written down isn't evidence

This is the belief the whole thing is built on.

Most first-round conversations leave behind a feeling and a few lines of notes. Ask three interviewers why a candidate did or did not progress and you will get three answers, none of which can be checked. The conversation happened, and then the record of it evaporated.

Unstructured interviews mostly measure how much the interviewer enjoyed the chat. That is not a slight on interviewers, it is what happens when the same person improvises a different conversation forty times and then compares the results. Structured interviewing is the best-evidenced idea in hiring and the least practised one, because being consistent at volume is genuinely tedious work for a human being.

It does not have to be a constraint any more. That is the opening.

What we're building

Four parts.

A ten-minute conversation instead of a form. You describe the role out loud to an agent that has already read up on your company. What comes out is a written brief and a scoring rubric naming the specific things a good candidate has to be able to do.

Candidates who hold their own profiles. Built from a CV plus one conversation covering what a CV cannot say: what someone actually wants next, what they are good at that is not on the page, what they would rather not do again. That profile belongs to them. We match it against your rubric and show you the reasoning rather than a percentage.

A structured first-round interview, in the candidate's language. The same questions, scored against the same criteria, in any of 43 languages, because someone thinking in their own language gives you a better read than someone translating in real time.

A report that cites its sources. Not a recommendation. Scores against your rubric, and under each score the moment in the interview it came from, quoted and timestamped. When you take someone to your own panel you can show your reasoning without making anybody watch an hour of video.

That last part is the one we care most about. The output is an audit trail by default, not because a regulation asked for one.

What it will cost, and when

An agency charges 20 to 30% of first-year base salary, payable when the hire starts. We charge 10%, on the same trigger.

No subscription. No posting fee. Nothing to approve before you begin. If the person leaves within 90 days, we replace them.

We are not framing that as a discount. Charging on outcome is the honest version of this business, because it puts our incentive where yours already is. Anything else asks you to pay for activity and hope the activity was any good.

The part that should make you pause

There is a fair objection to all of this, and it deserves a straight answer rather than a reframe.

Should an AI be interviewing people at all?

The honest scope is narrower than the question assumes. A structured interview replaces an unstructured phone screen. It does not replace a hiring decision, and we are not building it to. The decision, the rubric and the shortlist stay with you.

And nothing happens to a candidate without their say-so. Their profile is not public, it is not searchable, and it does not go to a company they have not agreed to. If a company passes, they hear about it. Those are not features we are adding for goodwill. Candidates are the supply this depends on, and a supply side that feels processed does not stay.

What we are actually competing with is not a careful human conversation. For most applicants right now it is a keyword filter and no conversation at all.

Where this goes

We will keep writing about it as we build: what structured assessment actually catches that a CV screen misses, what auditability in hiring should mean before somebody legislates it, and what recruiting ought to cost.

In the meantime, a genuine question, and the answer is useful to us. What does your first round look like today, and could you show somebody why you passed on your last five candidates?