Contact
The Candidate Pledge is a set of commitments to the people on the other side of the interview. Not aspirations; commitments.
Read those again and notice what they demand of us. Every one of them is a claim a candidate, a recruiter, or a regulator could hold us to. That is the point - accountability matters and we take it seriously.
Fair, explainable, ethical, responsible. A few years ago those words set a hiring tool apart. Today every vendor uses them, which means on their own they tell a buyer nothing. The words are easy but the architecture behind them is not.
Here is what makes ours true.
Hubert runs on two distinct layers. The conversation a candidate has feels human because it is designed to; the assessment that produces a score does not run on a probabilistic LLM. It runs on deterministic models. Same input, same output, full explainability.
That distinction is the whole game. With LLM-based tools, the same candidate answers can return different scores on different days, and the explanation is often a plausible-sounding narrative written after the score was decided. Hubert works the other way around. The reason a recruiter sees a score is the exact logic that produced it. Our explainability is faithful, not plausible. And because the models are deterministic, the same answers get the same score, every time. A candidate's result does not depend on which server processed their interview or the time of day they took it.
This is what we mean when we say fair. Not a value we assert, but a property you can verify: every candidate gets the same structured, competency-based interview, assessed against the same criteria, scored by the same locked models. Fair because it is consistent, and consistent because it is deterministic.
A pledge is only worth signing if candidates feel it.
In practice it means you are assessed on what you say about your skills and experience, not on how polished your CV is or where you went to school. It means that if you ask why, there is a real, specific answer, because explainability is built into the score rather than bolted on afterward. It means your interview data is treated as yours, processed and stored in the EU, and never used to train third-party models. And it means feedback, because being left to wonder is its own small unfairness.
The result is a 9/10 average candidate experience across deployments and candidates consistently rating us on Google 4.8*/5 with comments such as ‘To be perfectly honest, I've never had an interview like this before. It was fantastic, and the time allotted was a huge advantage. You could take your time to think about your answers and write them down. Then you could proofread them before sending them. Perfect, and please only do it this way in the future!’
The same design that protects candidates is what lets a recruiter stand behind a decision.
Because every score ties to a specific response and the full reasoning is on record, you get scored, auditable shortlists directly in your ATS; shortlists that are transparent and justifiable to stakeholders and regulators. The final hire or no-hire call always stays with your team. Hubert is a high-speed research assistant, not a decision-maker. Tt surfaces the candidates who best fit the criteria and a human clicks the button.
This is legally defensible by design, and it scales. At ManpowerGroup it has meant a 67% reduction in screening time for recruiters. As Michael Stull, Managing Director at ManpowerGroup UK, puts it: "We've reduced screening time by 67% for our recruiters." At NSS Group in the UK, skills-based screening with Hubert produced a 50% increase in hires from candidates who would never have passed traditional CV screening; talent that consistent assessment surfaced and CV filters had been throwing away. Across the platform, teams see up to 80% less screening time and 80% faster time-to-hire, in 30+ languages, with 5x greater accuracy than traditional methods.
We are glad there is a G2 Grid for AI Interview Agent Software, and that you will find Hubert on it. Not because of where a logo sits on a chart, but because the category being measured at all is good for buyers. The more this kind of software is compared in the open, the more the questions that matter get asked out loud: explainability, consistency, candidate experience. Those are the questions we built Hubert to answer.
If you want to see how the Pledge holds up in practice book a demo to see how Hubert works.