‘Complicated and interesting questions motivate us’

“There are a lot of ways to look at data and give it meaning. To us, the data means the banks, clients, and their financial behavior. We look back in time to study it, and then we use it to predict how clients will behave in the future. That way, we can provide loans that are reasonable and scrutinized, and that present responsible risks.”

A box of apples or oranges?

“Data from 20 years ago looks different from today’s data. There are other files and structures, names, currencies, etcetera. That’s where our craftsmanship comes in. We take the past and make it look like the present, then use it to predict the future. To do that, we have to know the business and the bank very well. We also need a sense for data. You can’t just initiate an SQL query and then go for coffee. So what do we do then? We take a close look at the data and ask ourselves: ‘Is this logical, is this the number of missings that I’m expecting? Does this order make sense?’ You always have to be alert with historical data. Just because it says apples on the box, doesn’t mean that it’s apples in the box. Sometimes it’s oranges, and if you’re having a really bad day it’s a tractor.”

Puzzling and scrutinizing

“Our offices often have puzzles laying around. Sudokus, crossword puzzles, you name it. That’s how a lot of the people in my team choose to relax. We’re all a bit geeky and nerdy, and to be honest that’s a real pleasure. It’s great to be surrounded by colleagues who think like you do. People who look at something and say: ‘Something’s not right about that’, and then spend half the day scrutinizing it to to find out exactly what the issue is and what caused it.”

“That analytical quality is also important in the work we do here: solving really tough problems, like dealing with the ever-changing regulations in the banking world. Interpreting 15-year-old data so that they are in line with current regulations is like solving a huge puzzle that someone has stolen a big handful of pieces from, and all of the colors are wrong. Those kinds of complicated and interesting questions are what motivate my team and me.”

Our most powerful tool

“Technology is racing forward in our field, but we’re conservative with innovations. Banks are highly regulated, and with good reason. We have to be 110 percent certain about what we’re doing, and the models that we use to do it. Right now, neural networks are on the rise. You can do some fantastic things with those powerful systems, but there’s one major problem; you don’t know how they do what they do, so you can’t always understand why it spits out a specific outcome. When we create models that are supposed to predict the financial risks of the bank and our clients, we can’t rely on those kinds of black boxes.”

“So we use simple software to solve data puzzles: SQL. We’re on the verge of rolling out a more complete programming environment, but to be honest, the most important and powerful tool is the brain of the person working with the data. We run all sorts of statistical tests on the data, but the real work is looking at the results afterwards. ‘Is that what we were expecting? Does it mean what we think it means?’ We write code, but it’s your brain that really solves the problem.”

Community of the like-minded

“It might not always look like it from the outside, but data analytics and engineering are team sports. Our work is always better when multiple brains are working on it. That’s why we work with the buddy system and in teams, so you never have to solve a problem entirely on your own. There’s always at least one other person to act as a sparring partner, and that makes our work better. The team supports you and challenges you. Your co-workers look at your work and ask questions about what you’ve done. And when you get stuck, they help you by thinking about a solution.”

“Many people in our team come straight out of college, and for the first time they find themselves in a community of the like-minded; people who do the same thing they do. Where they have older colleagues to learn from, and where they can teach younger people in turn. Where they can grow into their role and where they’re appreciated and celebrated – for things that outsiders might not even understand.”

Failing and having impact

“What gives me energy is when I see people working on their puzzles with passion and enjoyment, and I watch them come up with new ways to work. Something I’ve introduced is ‘fail fast’; it’s another way to look at and deal with failure. Try something, and if it fails … great! Then try something new. The importance of it is the understanding that failure isn’t bad. You’ve tried something, and you’ve learned from it – at the very least, that it doesn’t work. Great! Now pick yourself up, and try something new.”

“Another reason why I get out of bed in the morning, is the knowledge that my work has a major impact here. Together with my team, I ensure that our data modelers work with reliable information that reflects what really happened in the past. Our work produces models that determine the stability of the bank and of the banking system as a whole, and from there the stability of the economy of the Netherlands and Europe, and eventually the entire world. That impact sets the bar high. It’s only by working and learning together that our models of the future can actually ensure our future.”