Models, predictions, and explainable reasoning
Every decision Customer360 makes is grounded in versioned models with confidence, drift, and feature attribution.
Churn Predictor
Classificationv4.2.1Predict 30-day churn likelihood per customer
Engagement decay across email + app, combined with two unresolved support tickets and a falling NPS slope, are the strongest churn signals this week.
Propensity to Buy
Classificationv2.7.0Score likelihood to purchase a product line in next 14d
Customers showing repeat browsing of premium SKUs with at least one wishlist add convert at 4.1× the baseline rate.
Next-Best-Offer
Recommendationv3.1.4Recommend the most profitable, accepted offer per customer
Offer ranking optimises for blended margin × acceptance probability, suppressing offers a customer has rejected twice in 60 days.
Behavioural Segmentation
Clusteringv5.0.2Cluster customers into actionable behavioural segments
12 stable behavioural clusters detected. Two are emerging (luxury-seekers, value-hunters) and warrant new playbooks.
Customer Health Score
Regressionv1.9.3Composite 0–100 score reflecting relationship health
LLM-derived sentiment from chat + reviews now contributes 20% of the score, surfacing dissatisfaction before it becomes attrition.
Profile Narrator (LLM)
LLMv1.2.0Generate plain-language summaries and explanations
Grounded on canonical profile + last 90 days of events. Refuses to speculate beyond evidence; cites sources inline.