My Portfolio – Nangosyah Tom Wellard

A tiny end-to-end ML pipeline for Kiva Loans

This is a small, low stakes example of building a complete machine learning pipeline in Python, from pulling raw data off a public API all the way to a working app that…

Clean Cooking & Productive-Use Carbon Model

Emission reductions, carbon-credit revenue and Gold Standard cost estimation

Business Analytics

This report estimates the carbon performance and economics of a portfolio of clean-cooking and productive-use (PUE) assets i.e., improved cookstoves, electric pressure…

Does Pre-Processing Locally Actually Pay Off?

Business Analytics

In July 2026, INENSUS, a German mini-grid developer working across rural Africa, put out a short LinkedIn deck with a simple pitch: mini-grids grow faster when they are tied…

Running an A/B Test in Business

A practical walkthrough using a loan pre-approval experiment

Business Analytics

A team wants to change something, a form, a price, a feature, because they believe it will improve results. That belief is often reasonable, but it is still a guess. Rolling…

Building Spotify-Style Playlists with K-Means

AI / Machine Learning

Spotify’s catalogue exceeds 100 million tracks, which means no human can possibly go through all that music and curate playlists, so it begs the question how does the app…

Formula One Performance DNA (Japan Update)

Sports Analytics

After the first two races I published a performance tier analysis that used sector times, top speed, qualifying pace, and race pace to cluster the eleven 2026 teams into…

Mapping Formula One Performance Tiers with Hierarchical Clustering

Sports Analytics

My journey with Formula 1 dates back to secondary school. My siblings and I grew up sitting with our dad in the living room, watching these cars go around circuits on…

Kampala’s Transit Network

What a 2019/20 dataset reveals about how Uganda’s capital moves

Geo-Spatial Analytics

Public Policy Analytics

This analysis is based on a publicly released dataset of Kampala’s Greater Metropolitan Area (GKMA) paratransit network, collected by a consortium of urban transport…

Symptom Extraction Using LLMs on the Vlaams Supercomputer

AI / Machine Learning

Health Data Science

I completed a 3-month internship with the Data Science Institute (DSI), UHasselt, exploring how open-source large language models (LLMs) running on high-performance…

Is 4/20 Dangerous for Driving? A Look at U.S. Fatal Car Crashes

AI / Machine Learning

April 20th commonly known as “4/20” is an informal, globally recognized holiday celebrated by cannabis enthusiasts. In the United States in particular, the day features…

Exploring the Diagnostic and Prognostic Potential of OCT Data in Multiple Sclerosis

AI / Machine Learning

Health Data Science

Machine learning (ML) is transforming Multiple Sclerosis (MS) diagnosis by leveraging Optical Coherence Tomography (OCT) data to assess retinal changes linked to…

Model Based Clustering of High Dimensional Longitudinal Data

Applied Statistics

Health Data Science

Using Model based Methods for clustering longitudinal high dimensional data, we used mixture models to classify peptides with similar deuterium exchange profiles overtime.…

Exploring the relationship between microbial composition and sample type using correspondence analysis

Applied Statistics

Health Data Science

The microbiome is the collection of microorganisms that live together in a particular place (e.g. river, soil, skin, gut,etc). With modern technologies we can identify…