AI-Augmented Solutions Developer and Data Scientist with a portfolio of 34 deployed projects and a foundation built across 15+ years in Nigerian secondary school classrooms. I operate across three builder identities: EdTech Builder (16 EdTech projects including a full CBT exam system and 11 LMS platforms), DataTech Builder (7 end-to-end ML pipelines across 7 industries, 11 data tool simulators), and FaithTech Builder (gospel-driven digital tools under HMG Gospel).
My teaching background is my professional edge — it gives me the ability to explain complex models clearly, translate technical outputs into decisions stakeholders can act on, and build solutions that work for real people in real conditions. I am currently active in three simultaneous advanced programmes: DeepTech_Ready (DSN × 3MTT × Google.org, DSML Cohort 3), WorldQuant University Applied Data Science Lab, and Kodecamp ML Core Track — all while still teaching every week in a secondary school classroom.
- Full-stack CBT platform built on Android tablet with ₦0 budget — no laptop, zero cost to schools
- 7-col CSV upload, timed exams, per-student randomisation, anti-cheat detection, auto-marking, certificates, 4-tab analytics
- Stack: HTML5 · CSS3 · Vanilla JS · Supabase (PostgreSQL + Auth + RLS) · Vercel
- Ecosystem: CBT Question Bank Manager · Student Study Plan Generator · SchoolDocForge · TimetableCraft
- NLP pipeline: text cleaning → tokenisation → lemmatisation → TF-IDF (20K features) + VADER sentiment → XGBoost
- 12,999 real news articles · AUC 0.9393 · Accuracy 86.75% — highest AUC in portfolio · Formally assessed 3MTT Capstone
- Gradient Boosting · 108 GridSearchCV configurations · 22,750 employee records · joblib serialisation
- R² 0.855 · RMSE 0.072 · MAE 0.053 · Mental fatigue (r=0.878) drives 85.9% of model importance
- Bank Customer Churn (Binary Classification) — Gradient Boosting · 10,000 records · F1 0.609 · AUC 0.868 · Live ↗
- Insurance Claim Prediction — Random Forest + SMOTE + SHAP + FastAPI · 7,014 records · CV F1 ~0.79 · Live ↗
- Yakub Staff Promotion (HR Analytics) — Random Forest + GridSearchCV + SHAP · 38,312 records · AUC 0.891 · Live ↗
- Income Level Prediction (Financial Inclusion) — 5-model comparison + SMOTE · 48,842 UCI Census records · Live ↗
- SwiftChain Delivery Delay (Logistics · Multi-class) — 15,549 records · Documents data leakage fix transparently · Live ↗
- LMS: SQL Learn · Streamlit Learn · Excel Learn · Data Science · Data Analysis · PyDataFlow · VizLearn · ML Lab · Junior Class · Senior Class · QueryPilot
- Simulators: MS Excel · PgAdmin · MS SQL Server · MySQL Workbench · VS Code · Azure Data Studio · DuckDB · Power BI · Tableau · Azure · ChatLens
- Founded and direct four-subsidiary organisation: HMG Academy (virtual school), HMG Technologies, HMG Gospel, HMG Media
- Apply data-driven methods to track student performance; manage WAEC/NECO/UTME registration with 100% data accuracy
- Built and deployed CBT Pro — production-grade exam platform at zero cost to Nigerian schools
- Use data-informed teaching to identify student knowledge gaps; improve WAEC/NECO pass rates (SS1–SS3)
- Personalised lesson plans for IGCSE, WAEC, NECO, UTME, BECE students across Nigeria and internationally
- Led all academic and administrative operations as Principal; taught Science and Computer Studies