Meet the Team: Jonathan Harrison, our new data scientist

We’re very pleased to introduce Jonathan Harrison, a Data Scientist at EPCON who joined the team in July 2025. Jonathan works at the intersection of data science and product development, helping make EPCON’s data platform far more accessible, especially for users without technical backgrounds.

A core focus of Jonathan’s work is integrating agentic, natural-language tools into EPCON’s system, giving users an intuitive way to engage with complex data. No SQL required.

What does this mean in practice? Instead of navigating dashboards or writing queries, users can ask for specific metrics (“Please give me the three municipalities with the highest vaccine coverage for children under five”) or drill into a particular context (“What is the population density and HIV prevalence in the Nelson Mandela Bay district?”) in plain English. This makes it easy for anyone to extract meaningful information quickly from EPCON’s systems. 

Jonathan’s path into this work is rooted in a rigorous quantitative foundation. He completed an undergraduate degree in Actuarial Science, an honours degree in Statistics, worked as a data analyst, and then pursued a Master’s specialising in Statistics and Data Science. His toolkit spans machine learning, statistical and Bayesian modelling, and natural language processing.

For his Master’s thesis at the University of Cape Town, Jonathan analysed prediction methods for short-term coastal water level time series to help forecast flooding risk in San Francisco Bay. One key insight: advanced neural networks performed especially well in the extreme short term, but their accuracy dropped as the forecast horizon extended – a reminder that model choice must fit the real-world question.

Jonathan loves being able to solve problems that have a tangible impact. He made a deliberate move away from purely speculative applications of data science in order to work on challenges that affect people directly. “It’s exciting to find fresh solutions to real-world problems,” he says, “especially when the work has an effect on communities across the world.”

Working at the edge of fast-moving fields comes with pressure. Jonathan notes that AI evolves at both speed and breadth, and staying current requires constant learning. He’s watching the rapid rise of agentic assistants closely, but believes we’ll soon see these tools become more specialised and better defined as the technology matures.

Looking ahead, Jonathan expects major improvements in infectious disease prediction to come from better data pipelines as more systems digitise. This translates into near real-time inputs that would keep models relevant. On the modelling side, he sees growing momentum behind ensemble approaches, where multiple models are combined to improve robustness and accuracy, including hybrid stacks that blend statistical and machine learning methods.

Still, the biggest challenge for data scientists remains the same: data quality and collection. “Garbage in, garbage out,” he notes. To counteract this, EPCON draws on a wide range of external data sources, which vary by country.  

Outside of work, Jonathan keeps active. He enjoys running, padel, and indoor bouldering, and is planning a hiking trip in Baviaanskloof later this year. He also reads widely, from fiction to classical literature - most recently The Iliad, which he describes as “surprisingly enjoyable.”

We’re thrilled to have Jonathan on the team and look forward to the ways his work will help make EPCON’s data and insights more usable, intuitive, and widely accessible.