Dan Steffy, PhD, Gurobi Optimization

My path to industry

For the longest time, I thought I would spend my entire career in academia. Yet now, I’m having a great time in a mathematical position in industry. I hope that sharing my experiences and perspective will be of value to some of you as you think about charting your own career paths.

As an undergraduate mathematics major, I set my eyes on becoming a professor. I loved mathematics and teaching and was attracted to the idea of academic freedom and flexibility; it seemed like a clear choice. In graduate school, I learned more about some of the interesting opportunities for mathematicians in industry. Although I got the idea that I’d probably enjoy either academia or industry, I still stuck with pursuing the former.

As luck had it, after a short post-doc, I got a tenure-track position in a math department; a career dream come true! Did I enjoy it? In many ways, yes, I loved making a positive impact on students through mentorship and teaching; I enjoyed having the freedom to define my own research agenda, and I had some wonderful colleagues.

However, after about ten years in the profession, I started to have second thoughts about whether it was still the right fit for me. With young children at home, maintaining a good work-life balance was becoming a top priority. I loved teaching, but not the assessment part of it. My administrative workload seemed to keep going up. Much of my time was spent on responsibilities that I was less passionate about; time for my favorite part, research, was often carved out of nights and weekends. I took a step back and wondered if this was still the best fit.

I reached out to some of my mathematician friends working in industry and asked them: what kinds of problems were they working on? What did their workday look like? Were they having fun? Their answers were very informative, and I liked what I heard. Before I knew it, I had an exciting opportunity to join some friends and former collaborators at a company called Gurobi Optimization and decided to make the jump.

What do I do now?

I have now been at Gurobi for going on two years. We develop software to solve several classes of mathematical optimization problems, including mixed-integer linear programs. The solver can be applied to a wide range of problems that impact our daily lives: optimizing airline crew schedules, deciding when to turn on and off generators at power plants, deciding where to install and upgrade cell-phone towers, choosing assets for investment portfolios, and the list goes on. We have around 150 employees, including a large PhD-level technical staff to develop and support our product, which implements many highly sophisticated algorithms.

As part of the Gurobi Experts team, my primary responsibility is to work directly with our customers to help them get the most out of our product. Many of our users also have backgrounds in mathematics, operations research, or related fields. I use my knowledge of mathematical optimization to advise them on technical issues and deliver training. I help diagnose unexpected solver behavior, tune our algorithms to be more effective for specific problem types, help remediate numerical issues, and troubleshoot deployment problems. In addition to directly working with our customers, I also participate in academic conferences, write articles for our knowledge base, and have time and freedom to work on side projects of my own choosing, including contributing to our R&D efforts.

What do I enjoy the most about my job? I feel like I’ve always got an interesting problem or puzzle to think about. Some of them come to me from our customers and some I can define and pursue on my own. Another aspect of the job I enjoy is that every day I am working with and helping our users apply mathematical optimization to solve a wide range of impactful problems that have a positive impact on the world. It is a contrast to my academic research, which was usually at least a step away from real use cases. Our company also values and prioritizes employee work-life balance; we have a flexible work schedule, and everyone in the company is fully remote.

Skill transfer and differences from academia

The skills I developed and used in academia have transferred very directly into my new role. Of course, my specific technical knowledge of mathematical optimization and general problem-solving skills are important, but it is much more than that. One very valuable skill is communication. Years of experience teaching and writing research articles helped me develop very effective skills in communicating complex technical content in clear and understandable ways. As I interact with our customers and colleagues, this is highly appreciated.  

The ability to quickly learn new things is an essential skill for conducting academic research and is also highly valued in industry. Some of the tools and technology we are using didn’t even exist a few years ago, so having the interest and mindset to continuously learn and grow is important. I love learning and applying new skills, so this has been very fun for me.

With all the similarities, what is different? Although the types of skills I apply are similar to before, there are several differences in how I spend my time and how my work is done. Due to the nature of the work, I now spend more time than before thinking about the technical aspects of mathematical optimization, which I really enjoy. I also find that many things move faster in industry. With my customer cases, I often solve problems, close cases, and move on to the next challenge quickly. We do have the freedom to work on a variety of medium or longer-term projects, but even those have shorter timelines than a typical academic research project. The speed also applies to ideas about business processes and other things in the company. If we have a good idea, it can often be put into practice very quickly, without the need for the sort of long committee review and approval processes required for many changes made at a university (or perhaps a larger company).

Finally, I will mention that my current industrial role has a very collaborative, team-based, environment. At a university, faculty are defining and carrying out their own independent research programs. They can choose to collaborate with peers and work with students, but I feel that as part of a team in industry, I am working more closely with my team to achieve a set of shared goals. This leads to more direct collaboration with peers in our daily work. I really enjoy the collaborative environment I’m in now, but I see that both of these environments have their own appeal.

Closing thoughts

Beyond building your skills in mathematics, here are a few bits of advice I can share, based on my own experiences:

• Learn to program. My own focus on computational research gave me a skill set in programming and computation that made it much easier for me to find opportunities outside of academia. Applying mathematical ideas to real-world problems requires computation, and computation is done on computers! Thus, it is important to know how to program them. I’m not saying to stress out over becoming a master programmer, but a solid foundation in algorithms and programming, paired with a background in mathematics, can be very powerful. Even if your interests lie in pure mathematics and you plan to stay in academia, I believe there are many opportunities to leverage computers. If you avoid learning to program, I think you are leaving a lot of opportunities on the table.
• Keep an open mind. For some time in the earlier part of my academic career, I had a rather fixed idea in my mind that I would stay in academia forever. In those early years, a few interesting industrial opportunities did come my way, but I politely declined without giving them a chance. Although I’m not sure that I would have taken any at those times, I probably should have opened my mind a bit more, even if to confirm that I really was where I wanted to be. Later on, opening my mind to new opportunities was an essential step in making my most recent career move, with which I am very happy.
• Talk to people. I think that the best way to get a realistic understanding of what a job will look like in academia, industry, or government is to talk to people in those roles. Before taking a big step in your career, seek out people in your desired next role: What does their workday look like? What do they like about it? Why are they happy (or unhappy)? Don’t be afraid to reach out to learn about people’s experiences; many are happy to discuss this kind of topic. This information, together with knowing yourself, will help you choose a path that is right for you. In addition to talking to peers or potential future peers, it is also valuable to build relationships with people a few steps ahead of you and learn from their wisdom.

I hope you learned something interesting from my experiences, and I wish you all the best of luck in finding a career that is a great fit for you.