A Data Scientist who ships what he models.

I started out at a bank, building revenue-assurance dashboards and automating data pipelines across billing systems. That year taught me something I have never let go of: getting a model to work is the easy part, and the real distance is between a model that works in a notebook and a system that delivers a correct answer, every day, to someone who is about to make a decision with it.

Today I am a Data Scientist at AI Creative Solutions, where I take the messy, undocumented data organisations actually run on and turn it into models, analytics, and decision systems they can trust. In practice that means feature engineering on data that fights back, analytics pipelines reconciled against a human source of truth, and governed AI agents that are kept nowhere near a number they could invent. I build the production platforms around that work too, because a model only matters once it reaches the people making decisions.

Titles blur at this layer of the work. In any given week I am doing data engineering, machine learning, business intelligence, and software engineering. I call it data science because every layer exists for one outcome: a correct number in front of a person about to make a decision. The pipelines feed the models, the platforms and agents deliver them, and the BI makes them legible. I build the whole path because a model that never ships is a lab result.

I have a strong opinion about applied AI: the model is the easy part. You can swap it on a Tuesday and nobody notices. The value lives in everything around it, the data wrangling, the feature work, the evidence governance, and the analytics that make a result trustworthy enough to act on. That is the part I build, and it is the part most demos quietly skip.

Beyond client work, I research applied machine learning for African field conditions, compete and organise as Zimbabwe's Zindi Ambassador, and write for audiences across the continent. I care about the same thing in all of it: building things that hold up when they meet the messy reality they were made for.

Technical range

Python, PyTorch, TensorFlow, scikit-learn, YOLOv8, T-SQL, PostgreSQL, Power BI, MindStudio, React, TypeScript, Express, Drizzle ORM, SAP Business One (SQL Server), a PostgreSQL-based ERP, and Azure cloud services. I work across the full data science lifecycle, from data wrangling and modelling through analytics to production AI systems.

Education

BSc in Data Science and Informatics (Upper Second Class Honours, 2.1), University of Zimbabwe. Graduating August 2026.