Understanding the analysis
Scores explained
VooHr displays several numbers and signal labels that come from different sources. They are not interchangeable and do not provide a diagnosis or a definitive prediction about a person.
Two different score sources #
The employee directory's wellness score and derived attrition-risk percentage come from a rule-based function using available overtime, absence, performance-change, missed-deadline, and recent engagement-survey signals. Missing inputs do not apply their penalty. This is distinct from the conversation analysis, which is generated by a language model from transcript text.
The analysis may also contain burnout-index and attrition-risk values, and the workspace can display a psychological-safety score. Review the labels and source before comparing numbers. A score from one conversation is not the same thing as the directory's rule-based roll-up.
| Score | What it measures | Range | How to read it | Limits |
|---|---|---|---|---|
| Directory wellness | Rule-based employee signals. | 0-100 | Higher values are better. | Missing signals are not penalized; it is not a complete wellbeing assessment. |
| Directory attrition risk | Derived from directory wellness. | 0-100% | Calculated as 100 minus wellness score. | It is not a trained forecast of resignation. |
| Analysis burnout index | Model-generated from the submitted transcript. | [CONFIRM: deployed schema range] | Read with the source transcript and evidence. | Not a clinical measure; may be absent or incorrect. |
| Analysis safety score | Model-generated signal for one conversation. | [CONFIRM: deployed schema range] | Treat as a prompt to review the conversation. | Not a validated survey or workplace measurement. |
What a number cannot establish #
Scores compress context and can mislead when data is missing, stale, inaccurate, or not representative. The AI-generated indicators are probabilistic and may be wrong. The rule-based directory score is mechanical: it reflects configured thresholds, not a complete assessment of wellbeing or intent to leave.
Use the transcript or signal inputs and your own review. A blank or 'No data' state means an input or analysis result is unavailable; it does not mean zero risk, no concern, or a positive result.
Use with care #
Do not use a single indicator as the basis for hiring, promotion, discipline, termination, or another consequential decision. Check the underlying evidence, confirm data quality, and give the employee a fair opportunity to share context.
The linked pages describe the implementation and limitations of each measure in more detail.
Use this guidance thoughtfully
This page describes the current product workflow; it is not an HR policy or a promise about every deployment. Before acting on information about an employee, verify the source record, consider context the system cannot see, and use your organization's process. If a control behaves differently from this guide, check the current workspace UI and confirm the saved result rather than assuming an action completed.