Skip to content


Greet and AI with Gustav Eklund

We talked to Gustav Eklund, machine learning engineer at Modulai, about how he got into ML, his previous experiences, and what he thinks about life at Modulai. If you’re a student, take part of the valuable tips he shares by the end of the interview.

How and when did your interest in deep learning and ML arise?

When I joined IBM as a trainee, there was a lot of talk about IBM’s Watson and what they referred to as cognitive computing. I did not follow a specific ML program when I studied at KTH but a lot of the stuff I studied was applicable to deep learning and ML. I worked closely with the Watson team and gained a lot of experience on how to use ML in business. I realized machine learning is a versatile toolbox and can be used to solve a lot of challenges, and I have been working hands-on with it ever since.

What is your role at Modulai?

I am mainly a machine learning engineer, which means I design, develop, and productify ML solutions in various contexts and for various businesses. I am also involved in managing projects, bringing in new exciting projects, and helping others develop their skills within ML.

What kind of ML projects have you been involved in over the years?

I have done a lot of projects within the retail sector, doing everything from recommendation systems to estimating the cost of stockout, workforce planning, and automating project onboarding. For the last few years, I have been quite involved in building Medtech applications, for example towards diabetes, dental care, and rheumatoid arthritis. More recently I have been involved in some finance applications for credit risk as well as building deep learning models for object detection, primarily for use on mobile and on edge devices.

What is your key learning from the 10+ ML projects you’ve worked on since you joined?

I have learned a lot of things during my time at Modulai. The level of experience in our team is incredibly high, and the combined number of successful projects from members of our team is staggering. I have learned a lot of new applications of ML, but also key factors that distinguish a good project from a  great project. In many ways, it’s about really understanding the context of which the ML system is going to operate, and what is important in that area of business. To be able to combine deep industry knowledge with deep knowledge about ML really sets the foundation for high-value solutions.

Before joining Modulai, you worked at a larger and more broad IT company, as an ML engineer and AI advisor. What would you say are the most significant differences between your previous ML experiences and those you gained here?

The level of ML knowledge in Modulai is really high compared to other settings I have been in. We are together keeping a close watch on advances in ML research, applications, and tools that enable us to solve new problems and work more efficiently. Furthermore, we have a great attitude towards getting things done, constantly improving ourselves, and how we work through collaborative learning.

What is the best thing about your job?

I am very fortunate to work with something I am so passionate about! We are a high-performance team and have a lot of fun together, sharing the passion for ML. We get to leverage state-of-the-art ML to develop cool solutions – that directly impact organizations, and help people to live better lives. I am really proud of the things we have accomplished since starting Modulai.

What are your main recommendations to a student starting their career in machine learning and ML?

If the goal is to launch a career and gain knowledge in applied machine learning, the best advice I can give is to get your hands dirty. Start to build ML models on real-world data, do Kaggle competitions, get to know tools that are helpful when doing real-world projects. Education in machine learning gives a great foundation, but many times a real-world project is different from what projects are done in school. The data may be of lesser quality, more data prep is needed, the problems are less constrained. Being able to assess what tools in the ML toolbox are best suited, and being able to communicate ML in layman’s terms are very valuable, and come with experience.

I am very fortunate to work with something I am so passionate about! We are a high-performance team and have a lot of fun together, sharing the passion for ML.

– Gustav Eklund, Machine learning specialist