The honest version

One job description, read through one person's stated hiring philosophy. Not a generic rubric, and not an average of everyone.

That is the whole point. The same CV scores differently with different judges, because the things they say they care about genuinely differ. A score here means one thing: how well your CV matches how that particular person says they hire.

No. Nobody is hiring through this page today. It is a way to see where you stand for a role type that is exploding, and to find out what a specific kind of hiring manager would push back on.

If a real employer later posts a role here, that will be obvious on the page.

No, and we say so wherever a name appears. Each judge is a set of written instructions describing how a public figure has publicly said they hire. It is not the person, is not affiliated with them, is not endorsed by them, and no opinion here has been given by them.

Only a profile marked Claimed has been confirmed by the actual person. Right now almost none are. If you are one of these people and want yours changed or removed, there is a claim link on the profile and we will act on it.

Yes, and we refuse to pretend otherwise. Below 8 real candidates on a judge, we do not show you a position at all. You get your score and the word Pending, with an explanation that the field is still filling out.

We also removed the fake candidates we had originally seeded, because a rank against invented people is a lie, and we were about to let candidates share it publicly. Your score is real from day one. Your rank becomes real when there are enough people to rank against.

Partly, and we would rather say that than claim otherwise. Keyword stuffing works less well than on a normal ATS because the judge is scoring against an opinionated philosophy, not a keyword list, and several judges explicitly penalise buzzwords and unverifiable claims.

Three things make it harder. You get one score per judge per email, so you cannot resubmit twenty versions and keep the best. The interview is a live conversation that probes the specific gaps in your CV, and vague answers move your score down as well as up. And the model is told to reward concrete detail, numbers and named projects, and to discount anything that reads like it was written to impress.

None of that makes it ungameable. A determined person can write a better document. That is also true of every CV ever read by a human.

No. When you use that route we record it, and we tell the judge explicitly that it is reading an unverifiable self-portrait written by your own assistant, that it should reward concrete detail and honest weakness, and that it must not score you higher just because the writing is warm.

We do this because otherwise the two routes are not comparable, and a well-prompted AI would beat a real career.

We store the text of it, your name and your email so we can score you, place you in a ranking and let you come back to your own results.

The text is sent to our AI provider to produce the score and run the interview. It is not sold, not used to train anything, and never shown to other candidates.

Only if you choose to share them. Other candidates appear to you as a rank and a score, never a name, and you appear to them the same way.

If you press share, we create a separate public page carrying your first name, your score and your standing. That page has its own link, deliberately different from your private results link, so sharing it never gives anyone access to your results page or your interview. It is also set to stay out of search engines.

Email quentin@treesos.io from the address you applied with, or reply to any email we sent you. We remove your record, your CV text and your interview.

No. It is one opinionated read of one document. We make no guarantee to you and no guarantee to any employer about anyone who scores well here.

Treat the score as a signal and treat the gaps as the useful part. The gaps are usually the thing worth acting on.

It is consistent, which is not the same thing. The same judge applies the same stated philosophy to everyone, but that philosophy is itself opinionated, and an AI reading a CV carries whatever bias sits in its training and in the instructions we wrote.

So it is not a verdict on your worth as an engineer. It is a structured, repeatable opinion, and it is worth exactly what an opinion is worth.

Trees OS. The contact for anything on this page, including corrections, deletions and complaints, is quentin@treesos.io.