About Me

I am currently a Solutions Architect for Databricks, working closely with customers at the early stages of commercial engagement, helping them to solve their data and machine learning problems using the Databricks platform.

I am also the co-founder of Bodywork Machine Learning, the creators of Bodywork - an open-source MLOps framework used by machine learning engineers to automate the deployment and execution of model-training workloads and model-scoring services, on Kubernetes.

Prior to this, I have worked with technology companies large and small, developing solutions to their data engineering and machine learning problems: I was Chief Data Officer for LiveMore Capital, where I built a cloud-based data platform to support their credit analytics and modelling activities; I worked within Oracle’s Adaptive Intelligent Apps team, automating workflows used to train, deploy and monitor machine learning models embedded within Oracle’s cloud applications; and prior to all this I worked for Perfect Channel, a medium-sized FinTech scale-up and provider of B2B transactional platforms, where I was responsible for the research and development of trading algorithms, and machine learning services that enabled clients to derive value from trade-event data.

Before moving into the world of data science and machine learning I spent several years working as a ‘quant analyst’ within London’s financial services sector. I was primarily focused on pricing and modelling the credit risk on derivative trades. So I know a thing or two about stochastic processes, time-series data, Monte Carlo methods and financial mathematics.

Going back even further, I graduated from UCL with a PhD in Computational Neuroscience, that explored how machine-vision techniques can be used to better understand human visual cognition. I learnt a lot about statistical learning and neural networks - before any of this stuff was as powerful and popular as it is now. I also have a background in theoretical physics.

I am very hands-on and have developed mathematical software programs for over 20 years, touching many programming languages along the way: Python, Scala, R, C#, C++, VBA, Java and Fortran. Currently, I have a strong preference for developing everything in Python (albeit in a functional style) and deploying to Kubernetes.

When there’s spare time I ride and race bicycles, write code (for fun!), drink good coffee and listen to house, techno and electronica (with a soft-spot for UK rave from the early 90s).