Martin Stoffel

Published 9th May 2025

How did you become an Research Data Scientist?

Previously, I was a researcher working on evolutionary genomics and biostatistics. I have always been very interested in coding though. I developed a few open source libraries for the science community and so that’s how I became generally interested in the field, I would say. I did a PhD in molecular ecology and then went on to do a postdoc at the University of Edinburgh. But I was a little bit skeptical about the normal academic path. I often felt that papers aren’t as impactful as I would want them to be. The software I’ve written turned out to be much more popular than any of my papers. So I thought it would be a great idea to just do that full time. At that point, I didn’t know about RSEs at all. By chance I saw the Alan Turing advert for an RSE position and just applied. Only in the course of applying did I actually find out what the job was even about. That you work together with researchers, but you do the data science and software engineering work. 

When did you first hear the term RSE?

I think during my interview at the Alan Turing Institute.

What is your favourite thing about your work?

At the Turing, we’re trusted to work in fields in which we don’t have a lot of experience – learning on the go. That’s very different in Academia, where it’s often very difficult to venture into other fields. But in research engineering at the Turing, it’s fairly easy. If you’re interested in a project, there’s a good chance that you can get on it, no matter your past experience. I think that requires great trust from the Institute. And it’s a fantastic learning opportunity. I love that.

What’s your least favorite thing about your work?

That’s a good question. I’m still a researcher at heart. And I feel we have a lot of freedom in our job. But we rarely have the chance to come up with new projects ourselves. Most projects have been broadly fleshed out before we research engineers come and work on them, as they are usually collaborations with Universities or the Government, for example. I think coming up with projects from scratch is what I miss most.

Do you kind of find a way to do it in the end? Do you normally have to kind of tow the line for your stakeholders?

In our team, we’ve got 20% time for non-project work. If you really want to do something, you can suggest it and might be able to start the project, for example as a service area. That means that it usually has to be of broader use for the group or the Turing. 

What’s the most unexpected part about your role?

I mean, people told me when I applied that we can choose projects and our past experience doesn’t matter that much. But the project I’m on now, for example, within the Digital Twin Turing Research and Innovation Cluster, I really had no idea about any of those things. And I was still trusted to work on it. That was a positive surprise for me. In the Research Engineering Group at least, direct subject matter experience is less important than interest and willingness to learn quickly.  

Do you see yourself as an academic, a research software engineer, a technician, all of it? Something else? Or a mix?

I feel all of it is maybe the right answer. I enjoy all of it. That’s why I’m doing what I’m doing. 

What do you see as your most likely future career path from here? And what would be your ideal career path?

That’s a difficult one. I’m thinking about this at the moment. I guess there’s a clear career path in our group at the Turing. When you’re moving up the ladder, there’ll be more management and less coding and research though, and I’m not sure I’m ready for that. 

If were a Senior Research Data Scientist would you still spend part of your time doing technical work and part of time management?

That’s right. Seniors in our team still code a lot, but proportionally just spend more time with other things like management and project scoping.

In your view, how could RSE or research data scientists be better supported in their work? What do you need? What’s missing?

The main thing that comes to mind is the RSE profession becoming more well-known and established. One of the hard things about the job is that when you work with researchers, and researchers are not actually full-time programming, not building software, it’s very hard for them to know how long stuff takes, for example. And of course, good software takes a long time and optimally needs a couple of people to write and to review code and have feedback loops. You don’t see these practices in academia much, so they aren’t really valued. I feel like the best thing is for events like RSEcon to happen and for the profession to become more established to help researchers understand and value what RSEs are doing.

What advice then would you give to individuals looking to start a career in research software engineering or as a research data scientist?

I guess like many of us, when you were in science before, the best thing to do is to take the coding part seriously and think about going a little bit further than maybe the average researcher and try to put your code on GitHub, use version control and try to make it reproducible. Maybe try to make a little package which will be something useful for others. You know, just diving a little bit deeper into the nerdy stuff. I think that gives a very good grounding of actually transitioning into an RSE role later on.