about me.

My PhD was a deep dive into the art of modeling data. I started by building a robust Bayesian model to make sense of NMR spectra, and I finished by analyzing images of leftover feed to teach a robotic milker exactly how much to give each cow.

Across these different challenges, I discovered that the central problem in data science is always model selection. Sometimes, you have an overabundance of inadequate tools, forcing you to adapt or build something new. Other times, you have inadequate data, forcing you to choose tools that honestly reflect that uncertainty.

I’ve built my approach around solving this puzzle. If the right data exists, I find the right tool for the job. And if the right tool doesn’t exist, I build it.

So, I have decided to do both. If the data exists, I want to look for the most appropriate tool to work on it, and develop one where necessary.