Contact
When thousands of people apply, shortlisting is slow for recruiters and invisible for candidates. Recruiters skim CVs under time pressure, and most applicants never get a genuine look. Hubert interviews every applicant with the same structured, skills-based interview, at any hour, in 30+ languages, then ranks the results so recruiters can focus on the decisions that matter.
The outcome is a process that is faster, fairer, and legally defensible: up to 80% less screening time, 2–5x higher screening accuracy than traditional methods, and a 9/10 average candidate satisfaction score.
Hubert uses two distinct AI layers. The conversational layer conducts the interview itself; it feels like talking to a thoughtful person, which is one of the reasons why 96% of candidates complete their interviews on average. The assessment layer scores the responses using deterministic proprietary models, not probabilistic LLMs. This means that every score ties to a specific response, models are versioned and locked per hiring cycle, and the explanation a recruiter sees is the exact logic that produced the score.
The final decision always stays with the recruiter. Hubert screens and recommends, and in the end, humans decide. There are no automated rejections and a full audit trail records every score and action.
Hubert is deployed across staffing, retail, logistics, healthcare and care, business services, and the public sector. Basically anywhere where high application volumes create pressure to move fast without cutting corners.
ManpowerGroup reduced screening time for recruiters by 67%. Ambea, one of the Nordic's leading care provider, supports more than 100,000 applications a year in Sweden with a single central recruiter after cutting screening activity by 74%. Malmö Stad cut a recruitment cycle from six weeks to two days. OKQ8 reaches a shortlist in an average of 57 minutes. Aleris went live in five working days from signing.
Hubert was founded before the generative AI boom. While others raced to wrap LLMs in recruitment language, Hubert had already spent years building proprietary models grounded in structured interviewing science. That head start is the edge, and it shows up in two commitments.
First, deterministic assessment. LLM-based tools can give the same answers different scores on different days. Hubert gives the same answer the same score, every time, which means it's repeatable and auditable by design.
Second, the Candidate Pledge: a commitment to candidate rights signed at C-level, not buried in a footer. It holds that explainability is a human right, that a black-box LLM should never decide a candidate's future, and that every candidate deserves feedback.
Speed or fairness is a choice you no longer have to make; with Hubert they are the same outcome. See it for yourself! Book a demo and get a shortlist you can stand behind.
No. Hubert conducts full structured, competency-based interviews through chat and voice, then scores the responses against explicit criteria. The output is a ranked, explainable shortlist delivered into the recruiter's ATS, not a scripted Q&A.
Hubert integrates with more than 30 ATS platforms, including Teamtailor. Interviews are triggered from the ATS and scored shortlists are delivered back into it, so recruiters never leave their existing workflow.
Hubert is built on the same core principles of transparency, fairness, and accountability that underpin the EU AI Act: faithful explainability, human oversight of every decision, and deterministic, auditable scoring. It is fully GDPR compliant; all candidate data is processed and stored within the EU, with no third-party model training on candidate data.
The core platform delivers structured AI interviews. Recruiter Agent builds structured interview templates from a job description. Job Coach guides candidates to the right roles on career sites. Outreach Agent re-activates talent pools via voice or text. HubertDetect™ flags AI-generated candidate responses for recruiter review.