Available for opportunities · Lagos, Nigeria

Adewale
Samson Adeagbo

10+ years in Nigerian classrooms. 12 deployed ML and EdTech projects. I use data science, machine learning, and AI to solve real problems — and I built my flagship tool on an Android tablet with zero budget.

Add Your Photo
CBT Pro — Live with Students
Educator · Builder · Data Scientist

I Did Not Start in Tech.
I Started in a Classroom.

For over 10 years, I walked into Nigerian secondary school classrooms — at schools like His Marvellous Grace, Fredaks, Anchor Heights, Dave Model School, Lifeline, and many others across Lagos and Ogun State — teaching Mathematics, Further Mathematics, Physics, Chemistry, and Computer Science. From JSS1 all the way to SSS3. I tutored on Tuteria, Gate Academy, DoLessons, and Prepclass. Teaching gave me something no university course can replicate: the ability to break down complexity, read where understanding breaks, and build the bridge that gets someone from confusion to clarity. That instinct lives in every model I build today.

The day I stopped complaining about broken tools and started building replacements — everything changed. CBT.ng (CBT Pro) was not born in a co-working space. It was born out of frustration, a clear understanding of the problem, and one Android tablet. I wrote every line of HTML, CSS, and vanilla JavaScript in Acode editor on mobile. No laptop. No budget. No compromise. Today it is live, used by real students, backed by Supabase, and hosted on GitHub Pages. That project proved a belief I now live by: a clear thinker with the right AI tools is unstoppable in any domain.

My formal data science journey accelerated through the 3MTT Programme in 2025, and I have not slowed down since. I hold a B.Sc.(Ed) in Computer Science Education from Lagos State University (2023). I have deployed 12 projects across machine learning, NLP, HR analytics, financial inclusion, EdTech, and logistics. I am enrolled in WorldQuant University's Applied Data Science Lab. I am an IBM SkillsBuild Ambassador (Cohort 4). I identify as an AI-Augmented Solutions Developer — an educator-technologist who uses first-principles thinking and AI assistance to operate confidently across any technical domain on demand.

Yes, there are days the imposter syndrome is loud. But then I look at the 12 live apps, the students who sat exams on the platform I built, and the classroom problems I turned into deployed solutions — and I remember: this journey is real, and I am exactly where I am supposed to be.

"If you can think through a problem clearly and ask the right questions, you can find a path to a solution — even in territory you've never entered before."

10+
Years Teaching STEM
12
Projects Deployed
1
Live EdTech Platform
500+
Students Reached
₦0
CBT Pro Budget

Skills & Tools

No meaningless progress bars. Just the real tools I pick up, learn deeply, and use to ship things that work in the real world.

Data Science & Machine Learning

The core of my transition — classification, regression, explainability, and model serialisation. Used across all 12 deployed projects.

Python Pandas NumPy Scikit-learn SHAP Matplotlib Seaborn Imbalanced-learn (SMOTE) Joblib (Model Serialisation) XGBoost Gradient Boosting Random Forest TF-IDF / NLP Pipelines Streamlit
Data Analysis & Business Intelligence

10 years of teaching Maths means I genuinely understand the numbers — not just the tools. I use BI to tell stories, not fill dashboards.

SQL Power BI Tableau Excel (Advanced) Google Sheets (Advanced) Pivot Tables & Statistical Summaries EDA (Exploratory Data Analysis) Univariate & Bivariate Analysis
Web & EdTech Development

CBT Pro was built entirely in these technologies on an Android tablet. No framework. No shortcuts. Pure understanding from the ground up.

HTML5 CSS3 Vanilla JavaScript Supabase (PostgreSQL + Auth) Row-Level Security (RLS) GitHub Pages Cloudflare Pages Acode (Mobile IDE)
Dev Tools & Workflow

Version control, notebooks, and deployment — the full cycle from experiment to live app.

Git & GitHub Jupyter Notebook Google Colab VS Code SQLite Gemini API (AI Explanations) Anthropic Claude API Formspree
Currently Learning & Deepening

Always in motion. The classroom taught me that the moment you stop learning, you stop being useful.

Deep Learning (Neural Networks) WorldQuant University — Applied Data Science Lab Microsoft Azure AI Fundamentals (AI-900) DSN × Google DeepTech_Ready — DSML Track Cohort 3 IBM SkillsBuild — AI, Data & Cloud (Ambassador) Piano & Guitar (ongoing since 2013)

Projects

Every project here started with a real problem — a broken classroom, a fraudulent claim, a student who almost failed but nobody noticed in time. The classroom ones most of all.

Binary Classification · Insurance
Insurance Claim Prediction
Predict fraudulent or high-cost insurance claims before processing — enabling smarter, faster risk decisions and stopping revenue loss at the source.
Random Forest · SMOTE · SHAP · 7,014 records
PythonScikit-learnSHAPStreamlit
Binary Classification · HR Analytics
Yakub Staff Promotion Prediction
Remove bias from HR promotion reviews by flagging high-potential employees objectively — giving Yakub Trading Group data-backed decisions instead of gut-feel ones. AUC: 0.891.
Random Forest · F1=0.411 · AUC=0.891
PythonPandasScikit-learnStreamlit
Binary Classification · Banking
Bank Customer Churn Prediction
Identify customers at risk of leaving before they actually leave — enabling proactive retention campaigns that protect long-term revenue. F1=0.609, AUC=0.868.
Gradient Boosting · F1=0.609 · AUC=0.868
PythonPandasScikit-learnStreamlit
NLP · Text Classification
TruthLens — Fake News Detector
Automatically detect misinformation in news articles — because not every student or reader can fact-check at speed. My first NLP project. Accuracy: 86.75%, AUC: 0.9393.
XGBoost · TF-IDF · Acc=86.75% · AUC=0.9393
PythonNLPXGBoostStreamlit
Regression · Workplace Wellbeing
NeuroWell — Burnout Rate Predictor
Predict employee burnout rate before productivity collapses — giving organisations an early-warning system they can actually act on. R²=0.855, RMSE=0.072.
Gradient Boosting · R²=0.855 · RMSE=0.072
PythonPandasScikit-learnStreamlit
Multi-class Classification · Logistics
SwiftChain Delivery Delay Prediction
Predict delivery delay category before dispatch — preventing logistics revenue loss before it happens. Notable: documented a data leakage fix transparently in the notebook.
Gradient Boosting · Multiclass · Acc=62% · F1=0.579
PythonGradient BoostingScikit-learnStreamlit
Binary Classification · Financial Inclusion
Income Level Prediction
Predict income bracket to power financial inclusion and credit scoring — especially for underbanked populations in the Nigerian context. UCI Census dataset, 5-model comparison with SMOTE.
Random Forest · SMOTE · 5-Model Comparison
PythonPandasScikit-learnStreamlit
EdTech · Binary Classification
Student At-Risk Predictor
Identify students likely to fail before exam season — so teachers can intervene early, not after results come out. Built on synthetic Nigerian secondary school data. Single-student and CSV batch modes.
Random Forest · SHAP Explainability · EdTech
PythonScikit-learnSHAPStreamlit
EdTech · Data Analytics Dashboard
Student Performance Tracker v3
Track and visualise student performance over time with 4 dashboard tabs — including at-risk thresholds and topic/question analysis. Gives teachers a real data picture of who is improving and who needs help.
Streamlit · Plotly · 4 Tabs · EdTech Analytics
PythonPandasPlotlyStreamlit
View All Projects in Full Detail

Learning Journey

Most portfolios only show the destination. This one shows the road — including where it began, the detours, and the deliberate moves that brought me here.

April 2026
IBM SkillsBuild Ambassador — Cohort 4
Selected as an IBM SkillsBuild Ambassador, championing digital literacy, AI fluency, and tech skills across Nigerian communities. Representing HMG Concepts and the educator-technologist path.
2026
DSN × Microsoft Elevate AI Developers — AI-900 Preparation
Deep immersion in Microsoft Azure AI fundamentals — machine learning concepts, responsible AI, computer vision, NLP, and conversational AI services. Building toward the AI-900 certification.
2026
DSN × Google DeepTech_Ready — DSML Track, Cohort 3
Intensive, industry-grade data science and machine learning programme in partnership with Google's DeepMind team. Applied learning on real datasets with real evaluation benchmarks.
2024 – Ongoing
WorldQuant University — Applied Data Science Lab
One of the world's most respected free data science programmes. Project-based, rigorous, and international. Working through real-world datasets, building and evaluating real models, submitting graded work to a global cohort. This is where theory meets professional expectation.
2025
3MTT Programme — Data Science Track (The Spark)
This is where everything accelerated. The 3MTT programme did not give me technical skills from scratch — it gave a 10-year problem-solver a technical language. Within months, I had gone from zero deployed projects to a growing portfolio, applying data science directly to classroom and real-world problems I had observed for years. The spark became a fire.
2025
12 Projects Deployed — ML, NLP, EdTech, Analytics
From notebook to live Streamlit apps: 12 real-world projects deployed across insurance prediction, bank churn, fake news detection, HR analytics, delivery prediction, burnout forecasting, income classification, student performance tracking, and EdTech tooling. All live. All solving documented real problems. All with honest model evaluation including the failures.
2024
CBT Pro — First Live Deployment with Real Students
The moment everything felt real. A full CBT exam platform — teacher portal, student exam engine, auto-marking, result certificates, analytics, AI explanations — all built on an Android tablet using Acode editor. No laptop. No budget. Used with real secondary school students for the first time. Backed by Supabase. The idea that became a product.
2023
B.Sc.(Ed) Computer Science Education — Lagos State University
A degree that formalised thinking I had already been doing in practice for years. The intersection of pedagogy and technology — understanding not just how to build, but how to teach what you build to someone who has never seen it before. That skill transfers directly to how I document and explain my data science work today.
2015
Founded HMG Concepts — His Marvellous Grace Educational Consult
From tutoring individual students to building a full educational brand. HMG Concepts now covers exam registration consultancy, edutainment events, and the growing HMG Technologies arm that is home to CBT Pro and future EdTech products. This was the first time I understood what it meant to build something bigger than yourself.
2013
Piano & Music Certificate — Rotop Music Academy
Before the code, there was music. Piano and guitar are still part of who I am — and if you think music theory and data science are unrelated, try explaining both to a 15-year-old. Pattern recognition, structure, and rhythm are everywhere once you know to look for them.
2013 / 2014
First Classroom — Where Everything Began
Walking into a Nigerian secondary school classroom for the first time as a STEM tutor. Maths, Further Maths, Physics, Chemistry — first at His Marvellous Grace, then spreading to Fredaks, Anchor Heights, Dave Model School, and more. I did not know it then, but those classrooms were training me in the two most important skills any data scientist can have: asking the right questions, and caring about whether people actually understand.
View Full Journey & Certifications

YouTube — HMG Concepts

The classroom, extended. Whether you want to see how a live data science project gets built, understand machine learning concepts from first principles, or prep for WAEC exams — it is all here, free.

HMG Concepts
youtube.com/@hmgconcepts
From building CBT Pro live on-screen to WAEC prep and data science project walkthroughs — this channel is where I teach the way I wish I had been taught. No jargon for the sake of jargon. No gatekeeping. Just clear, first-principles explanation from someone who has spent 10+ years making difficult things make sense.
Subscribe Free
How I Built a Full CBT Exam Platform on an Android Tablet — No Laptop, No Budget
HMG Concepts · EdTech Build Walkthrough
Insurance Claim Prediction — From Raw Data to Live Streamlit App (End-to-End)
HMG Concepts · Data Science Project Walkthrough
WAEC Further Mathematics — Arithmetic Progressions Made Simple (Pattern Recognition Method)
HMG Concepts · WAEC Exam Prep

What You Will Find on This Channel

Data Science Project Walkthroughs End-to-end builds: EDA → model training → Streamlit deployment. No hand-waving. Every decision explained.
EdTech Builds In Public Watch CBT Pro being built from scratch — architecture decisions, bugs, fixes, and everything in between.
WAEC / JAMB Prep Lessons Maths, Further Maths, Physics — explained the way that makes students actually understand, not just memorise.

Certifications & Programmes

Every programme here was completed — not just enrolled in. Each one was a deliberate choice to build a specific capability, not to collect badges.

DSN × Microsoft
Elevate AI Developers Programme — AI-900 Preparation
2026
DSN × Google
DeepTech_Ready — Data Science & ML Track, Cohort 3
2026
IBM SkillsBuild
SkillsBuild Ambassador Cohort 4 — AI, Data & Cloud Learning Paths
2025–2026
DataCamp
Python Data Fundamentals — Pandas, NumPy, Visualisation
2025
3MTT Federal Programme
Digital Skills — Data Science Track (The Catalyst)
2025
Digital Skills Nigeria
AI Fluency Certification — Responsible AI & Practical Applications
2025
Open Visualization Academy
Generative AI in Data Analysis — Prompting for Insights
2025
CareerEX / Youthrive · Access Bank
Career Readiness & Digital Economy — Professional Transitions
2025
Digital Skillup Africa · ST@40
Technology for Africa — Data Analysis & Skills Track
2024–2025
Jobberman
Soft Skills Certification — Communication, Leadership & Collaboration
2024
Earlier Technical Training
Diploma in System Data Management, Desktop Publishing & Computer Engineering
2014
Rotop Music Academy
Piano & Music Theory Certificate — The Pattern-Recogniser's First Curriculum
2013

From Students, Colleagues & Collaborators

Over a decade in Nigerian classrooms, and now a growing circle of data colleagues — some voices from people who have seen the work up close.

"Mr Adewale does not just teach you how to solve the problem — he teaches you how to think. My Further Maths result went from a D to a B2 in one term. That doesn't happen by accident."

TF
Tolu F.
SSS3 Student · Lagos Secondary School

"I have seen a lot of free exam tools for schools — most are either broken or complicated. CBT Pro is different. It actually works, and my students took their mock exams on it without a single technical problem."

AO
Adeyemi O.
Fellow STEM Tutor · Lagos, Nigeria

"What stands out about Adewale's portfolio is not just that there are 12 projects — it is that each one is documented honestly. He includes the failures, the leakage fixes, the models that didn't work. That is the mark of a serious data scientist."

KN
Kunle N.
Data Science Colleague · 3MTT Programme

Have you worked with me, learned from my YouTube channel, or used CBT Pro in your school? I would genuinely love to hear from you — your experience matters and helps other schools and recruiters understand the impact of this work.

Share Your Experience

Let Us Build Something Together

Whether you are a recruiter, a school, a company with a data problem, or a collaborator with a shared vision — I want to hear from you.

I am actively looking for data science and ML roles, open to EdTech collaborations, and always interested in school partnerships for CBT Pro. Based in Lagos, Nigeria. Open to remote.

WhatsApp (Fastest Response)
+234 810 086 6322
Email
hmgconcepts@gmail.com
LinkedIn — Connect & Collaborate
adewalesamsonadeagbo
GitHub — See All Projects & Code
cssadewale

Send a Message