Attrition Prediction Engine reference profile.
Which customers are at risk of churning or canceling, and when?
Overview
Gives Customer Success teams advance warning to intervene before a customer churns rather than after.
Who owns this decision
Chief Customer Officer, VP Customer Success
What this Specialist Model actually does
Decision: Which customers are at risk of churning or canceling, and when?
The final taxonomy records 5 supporting or external components, but it doesn't promote an Oumi-owned specialist model for this engine.
Who uses this Specialist Model
- Chief Customer Officer
- VP Customer Success
- VP Revenue Operations
- Head of Retention
Where the specialist model sits
The final taxonomy records 5 supporting or external components, but it doesn't promote an Oumi-owned specialist model for this engine.
Feature store aggregating usage, billing, support, and survey data
Supporting component
How Oumi connects to your stack →
Retrieval over historical account notes and prior save-play outcomes
Supporting component
Alerting/workflow integration
Supporting component
Alerting/workflow integration with CS platforms
How Oumi connects to your stack →
Tabular/time-series churn-probability model (gradient boosting or similar)
Supporting component
How Oumi connects to your stack →
Specialist LLM
Supporting component
Specialist LLM for risk-driver explanation and CSM-facing summaries
How to evaluate and improve it
Evaluation
- False positive rate (flagged accounts that renew)
- Revenue-weighted prediction accuracy
- Intervention success rate on flagged accounts
Feedback signals
- Actual renewal or churn outcome
- CSM-logged intervention and its result
- Account health score override by a human reviewer
- Win-back or save-play success/failure
- Downgrade or contraction event
- Customer-reported reason for cancellation (exit survey/interview)


