Replacing an employee costs between six and nine months of their salary. The earlier you know someone is considering leaving, the more time you have to intervene. The Retention Risk Intelligence Engine monitors digital behaviour patterns that research has shown to correlate strongly with resignation preparation.

The engine analyses four categories of signals: job search activity (including incognito searches), resume preparation behaviour, AI assistant usage patterns that suggest job-seeking intent, and broader behavioural changes like declining engagement and after-hours job-related activity. Each signal contributes to a composite daily score per user.

When the engine detects critical phrases in AI prompts, such as resignation letter requests or interview preparation, it can automatically escalate the user’s risk level. Cross-signal correlation further amplifies scores when multiple indicators appear together, because an employee who is searching for jobs, updating their resume, and asking AI for resignation advice is far more likely to be planning a departure than someone who simply visited a job site once.