Hi it's Nathaniel. The text on this page was written by Claude AI from my prompts. I have reviewed and approve the content. Thanks!
About
Nathaniel D. Phillips, PhD
I'm a cognitive psychologist and data scientist who has spent the last fifteen years studying how people make decisions — and building tools to help them make better ones.
From mathematical models of the mind to the front of the classroom.
It started in academia. I earned my PhD at the University of Basel, where I built mathematical models of how people search for information, weigh options, and learn from mistakes under uncertainty. That work taught me something I still carry: the best solution to a complex problem is almost never more complexity.
During a post-doc at the University of Konstanz, I created FFTrees, an R package for building fast and transparent decision trees, which has since been used by over 100 teams of scientists. That was a turning point for me — seeing something I'd built take on a life of its own showed me how rewarding it is to make tools that others can use (I'll come back to open source and R later).
Along the way, I taught statistics and decision-making at every level — bachelor's through PhD — and ran data science workshops for industry professionals. Teaching became less of a side activity and more of a through-line. Helping people understand what they're doing and why has shaped every role I've taken since.
Finances pushed me out of academia. What I found in industry made me stay.
I wish I could say that the move from academia to industry was to work on real problems, but it wasn't. It was to pay off student loans, and only meant to last a year or two. But what started as a financial decision turned into something I didn't expect. Industry was full of great science, complex problems, and a diversity of people and perspectives I hadn't found in the lab.
At Roche, I got my first taste — building analytics tools for real-world evidence in oncology. At Flatiron Health, the scope grew. I led development of tools that transformed how teams worked with electronic health records and genomic data, and helped standardize practices across a large and growing team of statistical programmers. That role cemented something for me: the work I care most about sits at the intersection of rigorous science, practical tooling, and clear communication.
The settings change. The core challenge doesn't.
Since then, I co-founded a healthcare analytics consultancy that worked with multiple health tech and biopharma companies. That experience stretched me in ways I didn't anticipate. On any given day I might be writing code in the morning, mentoring a junior data scientist over lunch, managing a project timeline in the afternoon, and presenting findings to senior pharma executives by end of day. Scientific writer, data engineer, customer success, team lead — I learned to put on whatever hat the moment required.
The settings change. The core challenge doesn't: replace noise with clarity, and make sure the answer actually lands with the person who needs it.
Open code, open methods, open to scrutiny.
I believe the best analytics work is transparent. That conviction started with R, which was my first real love in programming. I wrote YaRrr! The Pirate's Guide to R, a book that's still being used by students and professionals around the world, and I taught it to anyone who would sit still long enough. I created FFTrees for building fast and transparent decision trees, and shared a number of other R packages for tasks ranging from creating survival models (iwillsurvive), helping students use R (yarrr), and simulating solutions to statistical riddles (prisonersriddle).
Watching industries like pharma gradually move from proprietary software to open-source languages has been one of the most rewarding shifts I've seen in my career. These days, especially with AI reshaping how we work, my toolkit has expanded well beyond R. I reach for whatever is right for the job. But R will always hold a special place in my heart.
What's Next: Looking for the Right Fit (AI Will Probably Be Involved)
The next chapter is about bringing it all together. I want to help people and organizations figure out how to use AI in ways that actually matter — not chasing hype, but solving real problems. I want to help analysts and scientists flourish in their roles, not just survive them. I want to be part of scientific discoveries that improve lives beyond the people in the room.
Let's Work Together
Whether you're looking for a collaborator, a scientific leader, or someone to help your team level up — I'd love to hear what you're working on.