Make the data traceable
Lineage, controls and documentation are part of delivering reliable analysis.
About
Stephen Ahiabah
A risk and data professional with an engineer's instinct for systems and a statistician's respect for uncertainty.
I studied Materials Science and Engineering at Manchester, where experimental work taught me to be precise about evidence and honest about limitations. Banking gave that discipline consequences: models, controls and data quality affect real portfolios, regulatory decisions and customers.
I enjoy the work that sits between specialties, where a model, a dataset and a business question have to agree.
The route here
Career so far
2013
Materials Science
University of Manchester. Experiments, error bars and engineering maths.
2017
Graduate analyst
NatWest. Data innovation, Coutts private banking, fraud.
2019
Quantitative analyst
Stress test challenge across NatWest Markets and Ulster Bank.
2023
Franchise lead
Stress testing for a £200bn UK retail portfolio.
2024
Model lead
JPMorgan Chase. IFRS 9 and CECL for a new business.
2026
Data engineer
The data that every model depends on.
How I work
Four working principles
Be clear about uncertainty
Output should state its assumptions, its limitations and the range of possible outcomes.
Automate repeated work
Automation should cut manual effort and leave a process that is easier to review and maintain.
Use domain knowledge
Models work best when the question is framed properly and the result is read in context.
Outside work
Football analytics and writing
I'm a lifelong Arsenal supporter and the person behind CannonIQ and Pitch IQ. What began as a way to learn football analytics became a long-running practice in modelling, software, visualisation and writing.
- Based
- London, United Kingdom
- Studied
- Materials Science & Engineering
- Day job
- Data engineering at JPMorgan Chase
- Side work
- CannonIQ and Pitch IQ
- Tools I reach for
- Python, SQL, PySpark, Git