Mpho Mafata

Data Scientist and quantitative researcher with 5+ years of experience integrating complex biomedical and research datasets using R, Python and SQL. Experienced in data pipelines, statistical modelling, machine learning and FAIR data practices.

Health Data Scientist - Universitätsklinikum Tübingen - Germany

(2025-01)

The requirement I assessed for this role was that I needed to create a holistic, agile, and FAIR project management approach to ensure efficiency in daily functions. I then assessed the tools used by the group (Microsoft, Windows devices) and optimised them to achieve the goal of efficiency. Using simple tools (R notebooks and msWord), I managed to:

• Automate manual, repetitive data wrangling tasks (cleaning and pre-processing) reducing manual processing time from 2 weeks to three days
• Build, automate, and manage in-house reference libraries using open-source databases (query REST APIs for JSON and XML data formats)
• Develop frameworks for routine data analysis pipelines, causal inference (correlation and regression analyses), exploratory and predictive machine learning statistics (PCA, PLS, RandomForest) and creating automated reporting templates (LaTeX and markdown). All these reduced project completion timelines from two months to three weeks

Post-doctoral Research Analyst - Centre for Research on Evaluation, Science, and Technology (CREST) | Stellenbosch University - South Africa

(2022-12 - 2025-01)

In the role of researcher and data analyst in scientometrics (research publication and patent data analysis) and database development (PATSTAT, USPTO's PatentsView, and Clarivate's Web of Science - WOS), I structured and implemented patent statistics and indicators using a combination of PostgreSQL, Python, and R scripts. My work, focussing on patent-based technology trends, resulted in two research publications on the development of patent indicators. I also used cross database and platform validations (e.g., BigQuery open data vs local server DBs) and used dashboarding (Google Looker, PowerBI, and Excel) to create insights tailored for specific users and accessibility to the different dashboards.

Post-doctoral Research Fellow - South African Grape and Wine Research Institute (SAGWRI) | Stellenbosch University - South Africa

(2021-05 - 2022-11)

In this role I used Statistical modelling for Oenology, Food and Consumer Science of different data types, including Chemistry (targeted and untargeted LCMS, GCMS, MS/MS, IMS, IR, NMR) and Sensory (qualitative and quantitative) to develop strategies for modelling Chemistry-Sensory data fusion.

• Junior Associate of National Institute for Theoretical and Computational Sciences (NITheCS), South Africa

PhD - Oenology - University of Stellenbosch (2017-03 - 2021-05)

MSc - Wine Biotechnology - University of Stellenbosch (2014-03 - 2017-02)

BSc (Honours) - Chemistry (major), Biochemistry (minor) - University of Cape Town (2009-01 - 2013-12)