KenyansaAmenya/kenya-lend-intelligence

# KenyaLend Intelligence KenyaLend Intelligence is a backend platform for Kenyan digital lenders that combines customer intelligence, churn prediction, credit scoring, statement processing, and loan decisioning in a single FastAPI service. The project is designed as an API-first foundation for lenders that want to combine internal loan and customer data with M-Pesa statements, bank statements, and external macroeconomic datasets. > **Project status:** Alpha (`0.1.0`). Several endpoints and integrations are scaffolded and documented for extension, while some workflows still use placeholder logic or TODO-backed implementations. ## Table of Contents - What This Project Does - Core Capabilities - Architecture at a Glance - Technology Stack - Repository Layout - Domain Model - API Surface - Machine Learning Components - External Data and Integrations - Configuration - Prerequisites - Quick Start: Local Python Environment - Quick Start: Docker Compose - Database Migrations - Authentication and Authorization - Running Tests and Quality Checks - Deployment - Operational Notes - Development Workflow - Security and Compliance Considerations - Roadmap - Troubleshooting - Contributing - License ## What This Project Does KenyaLend Intelligence helps digital lending teams make more informed customer and loan decisions by providing APIs for: 1. **Customer management** - Create, search, update, and inspect borrower profiles. 2. **Loan lifecycle operations** - Submit applications, approve or reject loans, and summarize portfolio state. 3. **Credit decisioning** - Generate credit scores, risk levels, affordability estimates, default probabilities, and loan recommendations. 4. **Churn prediction** - Identify customers at risk of inactivity or attrition and segment customer cohorts. 5. **Retention actions** - Generate recommendations and track intervention actions for at-risk customers. 6. **Statement ingestion** - Upload M-Pesa and bank …