Tendong Brain Nkengafac

I build end-to-end machine learning systems that turn data into deployed, monitorable products. My work spans modeling, explainability, APIs, dashboards, and cloud deployment.

For an intelligent clinical decision-support system, I built a Kedro pipeline that benchmarked five models and selected CatBoost at 0.845 ROC-AUC. I added SHAP and LIME explanations, then delivered a FastAPI, PostgreSQL, and React application with clinical audit logging.

I've also built production-style forecasting and MLOps platforms for bike rental demand and freight pricing, including MLflow tracking, delayed-feedback drift monitoring, Docker deployments, and AWS ECS Fargate infrastructure. My freight model reached 1.88% MAPE, 40% better than a rate-per-mile baseline.

At ANTIC, I built a full-stack cell-tower geolocation system with cached, rate-limited REST APIs, geospatial processing, reverse geocoding, and a map-based interface. I'm looking to grow as a software and ML engineer while contributing to serious, impactful technology.