Caregiver Churn Prediction for Workforce Stability and Retention

A homecare service agency managing a large pool of caregivers working across part-time and full-time shifts. High caregiver turnover was disrupting client continuity, increasing hiring costs, and impacting the agency’s reputation.

The Need:
The client required an intelligent solution to identify caregivers who were likely to leave, based on behavioral patterns, shift data, feedback, and attendance. They wanted to proactively improve retention through early intervention and better support.

Our Solution:
We built a churn prediction model using open-source AI tools (LightGBM, pandas, scikit-learn). The model analyzed attendance trends, shift consistency, engagement scores, client feedback, and other performance indicators. An internal dashboard alerts HR and supervisors when a caregiver’s churn risk crosses a set threshold.

Outcome Achieved:

• Identified at-risk caregivers with over 85% prediction accuracy
• Reduced caregiver attrition through early interventions and check-ins
• Enabled HR teams to personalize retention actions (bonuses, role changes, mentoring)
• Improved client satisfaction through caregiver continuity
• Lowered rehiring and retraining costs over a 3-month window