Maureen Ngahu

I've built classification, NLP, and recommendation models with F1-scores up to 0.91, combining Python, SQL, machine learning, and statistical analysis to turn complex data into business outcomes.

At ACTOM Kenya, I analyze design-error data, lead root-cause investigations, and automate engineering workflows and reporting. This work improved engineering quality and output by 25% and reduced design-package turnaround time from five days to one day.

Previously at Schneider Electric Kenya, I analyzed ERP and material-planning data for procurement, inventory forecasting, and project execution. I also built sensor-data pipelines for PLC-controlled manufacturing systems, enabling real-time monitoring, process analytics, and proactive fault detection.

My project work includes an NLP pipeline for classifying mathematical misconceptions at an F1-score of 0.91, an XGBoost biodiversity model at 0.89 F1, and a hybrid movie recommendation system. I bring more than eight years of operational analysis, process improvement, and cross-functional problem-solving to data science work.