Senior Data Analyst with 8 years of experience building ML systems, marketplace optimization tools, and decision engines across banking, food delivery, ride-hailing, and Q-commerce.
Most of my work has been about taking messy operational problems and turning them into systems that are production ready. ETA prediction for 137K daily orders, credit risk scoring for 891K bank customers, real-time stress management for delivery marketplaces, shift planning, pricing, churn prevention. The projects below are the ones I can share publicly.
| Project | What it does | Domain |
|---|---|---|
| ETA Prediction Model | 5-stage segmented ETA with Bayesian smoothing and OSRM calibration | Delivery Logistics |
| Credit Risk Scoring Engine | MLP neural network scoring 891K customers with limit determination | Banking |
| Demand Control System | Zone-level stress detection with progressive marketplace tightening | Marketplace Ops |
| Courier Shift Optimization | Demand-aligned shift scheduling, improved utilization 0.8 to 1.1 orders/hr | Operations Research |
| Vendor Churn Prevention | Predicts vendor inactivity from operational KPIs, reduced churn by ~62% | Q-Commerce |
| Customer Pricing Intelligence | Geospatial persona inference + willingness-to-pay scoring for pricing | Pricing / Geospatial |
| Smart Ride Assignment | Replaces nearest-driver dispatch with ML-driven fulfillment ranking | Mobility |
Mrsool - ETA prediction, demand control, courier shift optimization
Careem - Experimentation framework, device allocation strategy, funnel instrumentation
Bykea - Ride assignment, customer pricing intelligence, dynamic surge pricing
Foodpanda - Vendor churn prevention across 92 zones
Bank Alfalah - Credit risk scoring for digital lending
Python, SQL, Scikit-learn, BigQuery, OSRM, GeoPandas, Statsmodels