What sparked your initial interest in physics?

I was always interested in taking things apart to understand how they worked. I think my parents found it a bit stressful at times, so it was cool that physics was a class at school that was all about figuring out how stuff works. It probably helped that getting good grades in physics was easier for me than learning a foreign language. During my final year of my schooling in Germany, we covered special relativity and had a whole week of project work preparing a presentation on the topic for the rest of the school. I think this really cemented my interest in physics and that I wanted to study it at university. I thought that if physics could explain space–time, then what couldn’t it explain? I did both undergraduate and postgraduate physics degrees at the University of Manchester, UK, including a summer studentship at CERN in Switzerland and two years based at Fermilab in the US.

How did you get interested in software development and machine learning?

When I was 13 I saved up my birthday money to buy a computer – I started writing small programs (and playing games). Then I got interested in building websites, and ever since, I have been fascinated with the wide variety of things you can do if you start plugging bits of software together, discover how they work on the inside and then modify them to do something new. It is kind of like gaining a superpower, one that you can learn! My first experience with machine-learning was during my PhD at Fermilab. The task was to identify electrons in the D0 experiment with high purity and efficiency. With youthful naivety I thought we could simply combine all existing discriminative features in an ensemble of decision trees. When that turned out to not outperform existing methods my curiosity grew. After my PhD I got involved with an open-source project called “Scikit-learn”. It is the most widely used software package for machine-learning in the Python programming language. The realization that I, or anyone really, could work with some of the world’s experts on this topic was a great discovery for me.

Did you ever consider a permanent career in experimental particle physics at CERN?

For sure. After finishing my PhD, I moved to CERN to work as a research fellow on the LHCb experiment. Being able to work in that environment for a long time was certainly very attractive. However, my perception was that the high level of competition leads to a large amount of luck being involved when it comes to who, out of a large pool of very good candidates, eventually lands that dream role. With my skills and contacts, I felt I could make a more direct impact on scientific progress by focusing on creating the tools that researchers need and training them in their use. That’s why I decided to leave my postdoctoral position at the Federal Institute of Technology in Lausanne in Switzerland after two years and founded Wild Tree Tech.

What are some of the challenges of moving from academia to setting up Wild Tree Tech?

The biggest challenge was having the self-confidence to take that step into the dark. I was fortunate to already have a good number of contacts all around the world from my open-source work. However, I didn’t know anybody in Zurich, where I had just moved to. So I spent the first few months doing nothing but meeting new people and telling them what I now do, while not being quite sure yet what that was exactly – combine that with trying to land my first contracts and wondering “Am I doing this right?”. Having a network of friends and more experienced entrepreneurs who I could turn to was very important. Navigating the administrative jungle was stressful but less difficult than I thought it would be. The biggest lesson I learnt is that you can just ask people for help. Almost everyone is happy to help you, or recommend someone that can. Most people remember how it was when they started out.

The biggest lesson I learnt is that you can just ask people for help.

Tim Head

What are some of the projects you are currently working on at Wild Tree Tech, especially in the field of open-source data?

Wild Tree Tech is a small consultancy that creates and customizes open-source software tools for data scientists and researchers. Our aim is to improve the user experience of analysing data and our motto is “data done better”. We are currently working on Binder – an open-source project makes it much easier to share your software with others. By clicking on a single link, a user can try out someone else’s project from the comfort of their web browser. This is great for sharing research code with colleagues, teaching material with students or even using whole books that contain programming exercises with your readers. Researchers at CERN are also interested in using it for reproducible research purposes, open-data initiatives or outreach events. It is nice to still have a link with my former life. Another favourite project of mine is setting up data-analysis infrastructure for the Open Humans Foundation to make it easier for its users to analyse the data shared via the platform. I also teach hands-on courses for data scientists, programmers and academics who want to learn the ins and outs of machine-learning. It is a great feeling being able to, in just a few days, instil the idea that this is something anyone can learn and understand how powerful these tools are. I feel like I am passing on my computer-given superpowers to others.

How has your physics background been helpful in your work?

Being able to look at very complex problems, and break them down into small chunks that I know how to solve is invaluable. Years of research have given me the confidence that handling unforeseen and as-yet-unsolved challenges on a daily basis is not something that I need to be worried about. After all, that is the definition of research! As a consultant, people come to me for expert opinions and advice. Truth be told, sometimes you just need to say, “I don’t know, but I am sure I can find out for you.” That confidence comes from having done research for several years.

Any advice for today’s students?

Keep learning new things. By trying out new things you will meet new people who share your values and ideas about life. Work with them on (crazy) projects, because like the click-bait headlines on the Internet say “You won’t believe what happened next”. With all your skills in physics, maths, thinking-like-a-scientist and hands-on hardware experience you are pretty much unique, on a global scale. Be confident of the skills you have and humble about the things you don’t know. Also, if you aren’t enjoying what you are doing now, consider switching to something new. I have yet to meet someone who regretted actively making a change, but I know many who never tried and wonder “what if”. Very often the only way to find out if something will work is to try doing it.

• Enjoy the rest of the April 2018 issue of Physics World in our digital magazine or via the Physics World app for any iOS or Android smartphone or tablet. Membership of the Institute of Physics required