Data Scientist, ML & AI EngineerData Scientist, ML & AI Engineer

Key Achievements:

M-Pesa Real-Time Fraud Decisioning Engine: Architected real-time pipeline on 120K transactions; engineered cyclical/ratio features and deployed sub-second FastAPI decisioning endpoint (LightGBM/XGBoost) returning fraud verdicts before settlement. Telecom Churn Analytics Platform: Built end-to-end customer churn classification engine analyzing usage patterns and billing metrics across subscriber records; deployed live stakeholder dashboard and FastAPI scoring API for retention strategy. Relational Sports Analytics Database: Designed and implemented normalized relational PostgreSQL database from scratch, establishing table relationships, indexing strategies, and automated integrity constraints for complex queries. MajiCast Geospatial Water Analytics: Integrated multi-source satellite, chemical, and geospatial data into an automated prediction pipeline combining XGB

Key Achievements:

• Directed commercial data operations, implementing structured inventory tracking and financial reporting systems that optimized profitability. • Conducted trend and sales pattern analysis across product streams to guide inventory procurement forecasting and stock allocation. • Executed critical data recovery and database integrity restoration across corrupted storage hardware, salvaging business datasets with zero data loss. • Delivered digital public services (E-Citizen, HELB) for 100+ clients with 100% adherence to data accuracy, privacy, and compliance standards. • Diagnosed and resolved hardware/software faults across 200+ computing systems using systematic root-cause troubleshooting.

Key Achievements:

• Processed, reconciled, and audited 1.5M+ national social assistance payroll records in advanced Excel, ensuring 100% adherence to statutory disbursement schedules. • Audited data within CCTP-MIS database, resolving duplications, rectifying 1,000+ client records, and maintaining strict payment data integrity. • Executed scheduled database backups directly at the primary server room level, guaranteeing zero downtime and absolute data safety. • Investigated and closed audit discrepancies within quarterly deadlines, collaborating closely with departmental stakeholders. • Maintained network and workstation infrastructure to ensure uninterrupted operational delivery across units.