What it means
AI score normalization is the process of adjusting or interpreting AI-generated scores so they can be compared consistently across records, workflows, reviewers, teams, or time periods. In HR software, scoring may appear in candidate matching, assessment support, workforce analytics, skills inference, engagement signals, or performance-adjacent insights.
Why buyers should care
Scores can look precise while hiding differences in data quality, model behavior, cohort size, job context, or reviewer expectations. Normalization can help make scores easier to compare, but it can also obscure important context if buyers do not understand the method. Buyers should treat AI scores as decision support that requires explanation, review, and calibration.
Evaluation checks
Ask vendors how scores are created, normalized, calibrated, and updated. Review whether the product explains confidence, source data, missing fields, cohort limits, and known failure cases. Check whether users can see raw context, override outputs, and audit score changes. Normalized scores should not become automatic decisions without human review and documented workflow controls.
This glossary entry is buyer-oriented guidance, not legal, compliance, or financial advice.