Data Scientist Mikael Karppinen, who has dedicated himself to neural networks, machine learning and Deep Learning, realised that he could secure the finances of millions in the financial sector with the help of artificial intelligence. The enormous masses of data, professional data scientist community and modern tools convinced this analytics consultant to choose OP Financial Group as his employer.

The details of Data Scientist Mikael Karppinen’s work tasks in Financial Crime Prevention at OP Financial Group are so closely guarded that even many of his colleagues are unaware of them. He is one of the approximately 30 data scientists at the Group’s Financial Intelligence Centre of Excellence. Even though Mikael is, through his position, naturally bound by a relatively extensive confidentiality agreement, we can still shed some light on how he ended up in his mysterious job at the Group.

Mikael began his career in analytics in consultancy firms that focused on reporting-oriented data science, which was the norm at the time. Later, he moved on to a smaller analytics consultancy firm that emphasised edge analytics,or machine learning on a terminal device.

“I used 95% of my working hours to code algorithms to a terminal device. It was interesting and developed my competence, but in the long run, this engineering type of work wasn’t for me,” says Mikael.

After spending years in consultancy, Mikael found himself yearning to be able to develop long-term solutions in AI, predictive analytics and machine learning. Gradually, the thought of an inhouse position became so tempting that he decided to go ahead with the move.

“The sprints for modern AI and machine learning solutions are long. It is challenging for a consultant to be able to be a part of developing long-term solutions that you’d be able to implement in the business after piloting,” Mikael says.

Modern technologies and vast data capital make OP Financial Group an attractive playground for Data Scientists

Last autumn, Mikael came across the Group’s job advertisement for a data scientist in Helsinki. Mikael had a strong brand image of the Group as a technology company that is currently undergoing an extensive digital shift.

“The financial sector has always been of interest. and I thought I could have a real impact on the lives of many people in Finland through AI and machine learning in such a large company. There are only a few companies in Finland that have data capital like OP Financial Group,” says Mikael explaining his reasoning.

He decided to apply for the position of data scientist, and he got to meet the people at the Group’s Financial Intelligence Centre of Excellence. Discussions quickly demonstrated that analytics-related projects that interest Mikael have been more than just talk and are being continuously developed at a fast pace in the Group.

“The Group has the latest technologies in use that enable neural network solutions, for example. In consultancy projects, it sometimes turns out that there are no data tools available, and then you have to do some data plumbing as well. Here, all that was already in order,” Mikael says.

By Finnish standards, the future team would also have an unusually extensive spectrum of in-depth expertise in data science and colleagues with whom one could break down development challenges, and of course be able to share your own expertise with others. This package was enough to convince Mikael, and he eventually joined the Group.

From consultancy to an in-house position – “My job as a data scientist at OP Financial Group has been a unique opportunity”

In the first six months at the Group, Mikael has had the chance to develop an AI-related solution in Financial Crime Prevention. Now his work includes more than just coding, mathematical modelling or hyperparameter optimisation. It is also directly connected to business and creating value to the end user in the long run.

“My job as a data scientist at OP Financial Group has been a great experience and a unique opportunity. In consultancy work, I had gotten used to the fact that development projects in large organisations are often hindered by bureaucracy. I’m surprised that hasn’t happened here,” Mikael says.

If Mikael had not ended up at the Group, he might have considered an inhouse data scientist position in a smaller company. However, it would not have been the best option for Mikael.

“It would be very unfortunate to be the only data scientist in the firm, and have only been hired because there is a lot of hype around data science. In this case, most of your working hours would be spent on explaining why there should be investments in data and why infrastructure and availability of data should be put in order,” Mikael says.

The extensive data scientist community has become important to Mikael in his inhouse position. Common discussions and developing ideas have completely swept this data scientist off his feet.

“In addition to enjoying having helpful and qualified colleagues, the know-how of the community is continuously growing,” Mikael says.

Currently Mikael is contemplating how “MLOps”, which has become the focus of machine learning application development, can be promoted at the Group. MLOps, abbreviated from “Machine Learning Ops”, refers to using familiar methods from DevOpsista in machine learning. At the Group, this practically refers to smoothly introducing models into production by removing technical friction from the development process and continuously monitoring the performance of the models.

“I believe we have all the ingredients to be able to come up with solutions that can withstand global comparison,” Mikael says.