Deterministic by design
Hubert interviews every candidate the same way and scores every answer on deterministic models. Same input, same output, full explainability.
Two ways to score an interview
Most AI hiring tools score candidates with large language models. LLMs are excellent at language, but scoring is different: careers depend on it, and it demands something stricter.
Deterministic models
- ✓The same answer always gets the same score
- ✓Every score traceable to explicit criteria
- ✓Models versioned and locked for the hiring cycle
- ✓Bias you can test, measure, and correct
- ✓Documentation ready for audit
Probabilistic LLMs
- ✕The same answer can get different scores between runs
- ✕Reasoning you cannot fully inspect
- ✕Behavior drifts as the underlying model updates
- ✕Bias is hard to isolate and correct
- ✕Difficult to document for regulators
Two layers, one platform
The conversation uses AI to feel warm and human. The assessment runs on deterministic proprietary models, and the explanation a recruiter sees is the exact logic that produced the score, not a story written after the fact. You do not have to choose between a human experience and an auditable one.
From answer to auditable score
Structured interview
Every candidate gets the same competency-based questions, by text or voice, in 30+ languages, around the clock.
Criteria-based scoring
Answers are scored on behavioral anchors in the STAR framework, with explicit weights recruiters can see and set.
Faithful explanations
The explanation shown to a recruiter is the same logic that generated the score; nothing is reconstructed after the fact.
The recruiter decides
Hubert surfaces a ranked shortlist with a full audit trail. No automated rejections; the final call always stays human.
Six principles of Responsible AI
These are the six pillars Hubert's assessment models and products are designed against to ensure responsible AI in hiring in everything we do.
- ExplainabilityTransparent explanations for every AI score. No 'black-boxes'.
- QualityScientifically validated assessments that predict real-world performance.
- ConsistencySame input, same output. Deterministic scoring.
- SecurityData privacy and data minimization by design.
- FairnessEqual opportunity for all candidates, regardless of protected characteristics.
- Human oversightDecision remains firmly with the recruiter.
Legally defensible by design
The EU AI Act classifies AI used in recruitment as high risk, with requirements for transparency, human oversight, and accuracy. Hubert is built on those principles by design, not retrofitted: every shortlisting decision is auditable, every assessment criterion is documented, and every candidate can understand how they were evaluated.
Frequently asked questions
What is a deterministic AI model in hiring?+
How is deterministic scoring different from LLM scoring?+
Does Hubert use AI for the interview conversation?+
Who makes the final hiring decision?+
See deterministic scoring in action
A 30-minute walkthrough with our team, tailored to your roles, volumes, and ATS.
Get a personalized demo
