Data scientist working at the intersection of health and finance. Turning messy data into decisions that matter.
"I don't just build models, I ship them.All creative ideas come from my midnight rabbit holes"
I build intelligent tools that translate complex, inaccessible information into clarity for real people , sparked by real frustrations I've actually lived through.Got a confusing medical scheme rejection letter? Been there. Stared at a credit report that reads like ancient scripture? Also been there. Built a tool for that too.
Most of my projects were born at 2am down a rabbit hole that started with one question and ended with a GitHub repo. That's just how this works.
What sets my work apart is that I don't stop at the notebook. I build, deploy, and deliver things people can actually use — whether that's a live ML application or a business insight that drives a real decision.
I'm equally comfortable doing deep data analysis and building end-to-end machine learning pipelines, and I'm always looking for the overlap between the two.
A deployed ML application that classifies cancer tissue samples — lung cancer tested against benign and colon cancer. Built to explore how machine learning can assist pathologists in diagnostic workflows. Anyone can use it, right now, in the browser.
.png)
An end-to-end analytics project on 100k+ orders from the Olist dataset, quantifying how delivery delays drive a 20–30% drop in customer satisfaction. Includes a churn prediction engine and hypothesis-tested seller performance analysis, delivered as an interactive Streamlit app and executive report
A credit risk model built on 2.9M behavioural transaction records to assess default risk for 500+ informal businesses. Features SHAP-based explainability for transparent model outputs and a Power BI dashboard translating predictions into actionable insights for non-technical stakeholders.
17 years of National Credit Regulator data (71 quarters) transformed into a production MySQL database with automated ETL pipelines. Features a deployed Plotly Dash dashboard and original research proving the lack of sustained recovery in the SA credit market since 2007 — with direct implications for lending strategy and financial inclusion policy.
.png)
Open to data analyst and data science roles, collaborations, and opportunities where data can make a genuine difference — especially in health and financial services.
See My Work