Advanced AI Capabilities
Built on your own data. Advanced models are rolling out as the product matures.
Churn Prediction
Survival curves + churn risk signals (beta)
RetentionLens computes Kaplan-Meier survival curves and hazard rates from your billing and lifecycle events, then surfaces explainable risk signals. Scores improve as your historical data volume grows; more advanced models are under active development.
Revenue Forecasting
Revenue dynamics + projections (in development)
Current release focuses on revenue dynamics (NRR/GRR, expansion, contraction, logo churn). Statistical forecasting with confidence intervals is under active development.
Customer Intelligence
AI-powered segmentation and health scoring
Automatically segment customers based on behavior patterns, predict lifetime value, and identify expansion opportunities with AI-driven customer intelligence.
Strategic Insights
AI-generated business recommendations
Get actionable business insights generated by AI that analyzes your data patterns, market trends, and industry benchmarks to recommend strategic actions.
Hybrid analytics + ML architecture
Next.js powers the UI and lightweight APIs; heavier compute is moving into a separate Python ML service that writes results back to Supabase.
Enterprise AI Infrastructure
Machine Learning Models
Survival analysis (Kaplan-Meier) is available today. Advanced models (Cox, ETS/ARIMA, clustering) are under active development with org-scoped versioning.
Continuous Learning
Scheduled batch jobs and a model registry (planned) will support retraining, evaluation, promotion, and rollback per organization.
Real-time Processing
The UI reads precomputed predictions and time series from the database for fast dashboards, while background jobs handle heavier processing.
Data & governance
Models are trained and evaluated on your own historical data where volume supports it, with safe fallbacks when data is thin.