Who I Am
I am a Data Scientist and BI Specialist with over a decade of academic and professional experience spanning machine learning, statistical modelling, and business intelligence across three countries and three continents. I believe data only matters when it changes something, and I have spent my career making sure it does.
Born and raised in Nigeria, I developed a rigorous foundation in statistics at the University of Ilorin, earning both a first-class Bachelor of Science and a Distinction-grade Master of Science in Statistics. Driven by a desire to operate at the frontier of applied data science, I pursued a second postgraduate degree in Data Analytics at Aston University in Birmingham, UK, graduating with Merit in 2023.
My professional journey began at Nigeria Inter-Bank Settlement System (NIBSS) in Lagos, where I spent over four years building production ML systems for one of Africa's most critical financial infrastructure providers. A deep neural network I designed and deployed achieved 97.84% annual forecast accuracy on product volume prediction, enabling 100% service uptime year-round and eliminating 52% of avoidable resource expenditure. I also led a full organisation-wide BI transformation using Power BI and Tableau that improved stakeholder report acceptance by 95%.
Since relocating to Calgary in 2022, I have worked as a Data Scientist at Canadian Blood Services, where I build ML systems that directly support the national blood supply chain. My work spans NLP-powered clinical data translation, real-time diagnostic and predictive portals, Monte Carlo simulation frameworks, Mixed Integer Linear Programming for strategic clinic placement, and supervised ML models that increased donor appointment attendance by 20%. These are not research projects: they are production systems used daily to ensure Canadians have access to the blood products they need.
What sets me apart is the combination of statistical depth and engineering discipline. I do not just build models; I build systems that work reliably in production, communicate uncertainty clearly, and translate directly into decisions that stakeholders can act on. I thrive at the intersection of rigorous analytical thinking and practical, measurable business impact.
02 Skills
A decade of applied data science across healthcare, banking, and national institutions. From statistical theory to production ML systems and enterprise BI platforms.
03 Portfolio
An end-to-end ML portal that translates messy clinical abbreviations into standardised CIHI codes using TF-IDF and K-Nearest Neighbours, powering a Random Forest model that predicts hospital blood demand with quantified confidence intervals to eliminate stock-outs and reduce waste.
View Details Internal deployment · CBS proprietary systemA real-time diagnostic system that surfaces appointment collection shortfalls across clinic networks, pulling live data from Azure Synapse and classifying each clinic as Normal, Warning, or Critical to expose root drivers instantly.
View Details Internal deployment · CBS proprietary systemMonte Carlo simulation (2,000 runs per clinic) with Damped Holt-Winters forecasting and bootstrap confidence bands to project future shortfall risk, feeding a proactive CTA framework that prescribes interventions before problems arise.
View Details Internal deployment · CBS proprietary systemA Mixed Integer Linear Programming model that uses demographic, geographic, and historical donation data to forecast active donor populations within specific Forward Sortation Areas, identifying underutilised donor potential to guide strategic clinic placement decisions.
View Details Internal deployment · CBS proprietary systemA comprehensive seasonality analysis across all variables in the Donor Appointments dataset to identify seasonal trends, peak and off-peak patterns, and cyclical drivers. Findings directly inform appointment scheduling efficiency and collection targets.
View Details Internal deployment · CBS proprietary systemEnterprise Power BI dashboards providing deep insight into Canadian insolvency trends, segmented by province, industry, and insolvency type, serving both operational planning and senior executive reporting.
View Live DashboardBig financial data distilled into a dynamic Power BI dashboard driving strategic decisions for Nigeria's financial sector players.
View DetailsPower BI dashboard showcasing strides in including Nigeria's underbanked population within the formal financial system.
View DetailsR survival analysis modelling time-to-employment for graduates, revealing structural barriers in Nigeria's labour market.
Read on MediumLogistic regression on clinical records from University of Ilorin Teaching Hospital to identify patient outcome factors.
Read on MediumPython CLI application computing descriptive statistics via Pandas and NumPy over US bikeshare datasets with interactive filtering.
GitHub RepoDeep SQL investigation of an online DVD-rental database using multi-table queries, window functions, and visualised outputs.
GitHub Repo04 In the Lab
Intelligent Emergency Care Navigation for Canada
Canada has some of the longest emergency room wait times in the developed world. Qura uses ML trained on hospital volume data, time-of-day patterns, and real-time occupancy signals to recommend the optimal care pathway for your situation: ER, walk-in, telehealth, or pharmacy. It tells you where to go, how long you will wait, and why, before you leave home.
Hyperlocal Winter Road Safety Intelligence
Every Canadian winter, thousands of accidents happen on roads that were passable thirty minutes earlier. GlazeIQ combines Environment Canada weather data, IoT road sensors, historical accident records, and crowdsourced driver reports to generate street-level ice risk predictions updated every 15 minutes. A spatial ML model gives municipal fleet operators, school boards, and commuters a genuinely useful risk map, not just a weather forecast.
AI Settlement Navigator for Newcomers to Canada
Canada welcomes over 500,000 immigrants annually. Settlement services are fragmented, generic, and often inaccessible. Berth ingests a newcomer's skills, background, language, and goals, then delivers personalised recommendations for jobs, neighbourhoods, and community networks. Unlike static government resources, Berth learns from real outcome data to improve its recommendations over time.
05 Writing
I write about data science, healthcare analytics, and the craft of turning messy data into decisions that matter.
How a MILP model using demographic and geographic data turned strategic clinic site decisions from intuition into rigorous science.
How NLP, Random Forest, and a deep respect for clinical messiness built a blood demand forecasting system that works in production.
Running thousands of simulations per clinic to forecast shortfall risk taught me more about uncertainty than any textbook ever could.
Abbreviations, typos, inconsistent coding. Clinical data messiness has real operational costs. Here is how TF-IDF and fuzzy matching clean it up at scale.
Most BI dashboards are technically correct and completely useless. The gap between ignored and indispensable is rarely about the data.
Three degrees, two continents, one pattern. The skills that transfer between contexts and the assumptions that absolutely do not.
A survival analysis reveals who finds jobs quickly, who waits longest, and what factors genuinely make the difference.
Age, not gender or length of stay, emerges as the primary predictor. Lessons from 320 patients at a Nigerian teaching hospital.
06 Experience and Feedback
The NLP blood demand portal Moses built is unlike anything I have seen in healthcare analytics. It took a problem we had lived with for years and made it disappear. The confidence intervals gave our inventory team something they had never had before: a number they could actually trust.
The Diagnostic and Predictive portals transformed how we operate. What used to take days of manual investigation now happens automatically, in real time. The CTA framework has shifted our whole approach from reactive to preventive.
The MILP clinic placement model gave us a rigorous, data-driven framework for decisions we previously made on intuition alone. The FSA-level analysis surfaced donor potential we had simply not seen before.
Moses brought a level of statistical rigour to our seasonality work that immediately elevated the quality of our planning conversations. The insights now directly shape our appointment scheduling targets every quarter.
The deep neural network Moses built at NIBSS achieved 97.84% forecast accuracy and contributed directly to 100% service uptime. What impressed me most was how he translated a complex technical solution into clear business outcomes that every stakeholder could understand.
Moses delivered a Power BI solution of exceptional depth. Complex national insolvency data that once lived in spreadsheets is now a living, interactive dashboard that our policy team uses weekly. The executive layer is exactly what our leadership needed.
Working with Moses on donor experience analytics was one of the more productive partnerships I have been part of. He combines rare technical depth with genuine curiosity about the business problem. The attendance prediction model changed how we plan collections.
The financial dashboards Moses built told a story with data that no spreadsheet ever could. The clarity of insight enabled real policy conversations that previously would have required weeks of report preparation.
The NLP blood demand portal Moses built is unlike anything I have seen in healthcare analytics. It took a problem we had lived with for years and made it disappear. The confidence intervals gave our inventory team something they had never had before: a number they could actually trust.
The Diagnostic and Predictive portals transformed how we operate. What used to take days of manual investigation now happens automatically, in real time. The CTA framework has shifted our whole approach from reactive to preventive.
The MILP clinic placement model gave us a rigorous, data-driven framework for decisions we previously made on intuition alone. The FSA-level analysis surfaced donor potential we had simply not seen before.
Moses brought a level of statistical rigour to our seasonality work that immediately elevated the quality of our planning conversations. The insights now directly shape our appointment scheduling targets every quarter.
Working with Moses on donor experience analytics was one of the more productive partnerships I have been part of. He combines rare technical depth with genuine curiosity about the business problem. The attendance prediction model changed how we plan collections.
Moses delivered a Power BI solution of exceptional depth. Complex national insolvency data that once lived in spreadsheets is now a living, interactive dashboard that our policy team uses weekly.
07 Credentials
Master of Science in Data Analytics
MeritMaster of Science in Statistics
DistinctionBachelor of Science in Statistics
First Class HonoursFull work history, certifications, and references available on request or via CV download.
Download CV / Resume