The oracle in Kastrup
Nikolaj Settnes helps PostNord see into the future. “A good algorithm can be delectable, maybe even a bit sexy.”
Text: Michael Kirkeby Photo: Freddy Billqvist
Nikolaj Settnes creates intelligent tools that predict how many parcels PostNord is expected to handle several months into the future. However, he is less interested in having total control over his own future.
“Since I was little, I’ve gotten used to taking one day at a time. Carpe diem thinking may sound romantic, but to me it is something completely concrete,” says Nikolaj.
He was in sixth grade when he got the news that his future was not like everyone else’s.
“Muscular dystrophy is caused by a genetic defect. I was just unlucky.”
When Nikolaj turns on his computer each morning, the goal is to have another good day at work. He is one of four highly specialized members of PostNord Denmark’s Business Intelligence development team. His workplace is on the top floor of the office complex on Hedegårdsvej in the Kastrup suburb of Copenhagen. The team are working to develop the digital tools and applications that will make life and work easier for their colleagues in the rest of PostNord.
“I’ve decided to turn on my computer with a ‘Yes! What are we going to do today?’ instead of a tired sigh. Life is too short for that. In my case, perhaps even shorter than for many others.”
Nikolaj was able to walk until he started university. He played guitar and sang in a band until just a few years ago. Today, he is 31 years old and prefers to focus on what he can do now – not what he can no longer do, or might not be able to do in a few years’ time. That doesn’t interest him very much.
It is a completely different story when it comes to his work tasks. Here, the IT tools that Nikolaj develops need to predict the future as accurately as possible. How many parcels can PostNord expect to handle in three months? How many people need to be called in to work at that time? If the forecast is off by more than five percent, it is not good enough. Nikolaj and his colleagues must then adjust the algorithm and make the tool better.
“In principle, PostNord Predict – or PNP as we like to call the tool – can always be made better. Our attitude is that the best can always be made better and ‘pushed a little further’ as we usually say. It’s a great professional satisfaction that I don’t want to be without. A good algorithm or code can be delectable, maybe even a bit sexy.”
PNP’s ability to predict parcel quantities is based on what is known as Big Data. The algorithm in PNP continuously registers and analyzes millions of pieces of data coming from different sources.
“There are many variables that affect parcel volumes,” Nikolaj says. “What do consumption patterns look like right now? When are wages, child allowance and holiday pay paid out? What was the period like last year? How many shopping days are left until Christmas? What does the vacation situation look like in other countries where Danes shop online? How can these and other factors heighten each other, creating the effect 1+1 = 3? It’s a huge puzzle that is constantly changing.”
Yet the forecast will never be one hundred percent correct. Reality does what it wants to do. As the saying goes “It's tough to make predictions, especially about the future.” The weather is never completely predictable, and there are sometimes phenomena like the coronavirus that turn everything upside down.
For Nikolaj, the pandemic meant that many months went by where he barely saw anyone other than his girlfriend Michelle, his regular assistants, and little Molly, his lively Bichon Havanais dog who happily runs around the apartment in Vallensbæk Strand, south of Copenhagen.
“I really don’t want to catch the coronavirus, so many months passed before I started meeting up with my family again – we met outdoors, maintained plenty of distance, and used lots of hand sanitizer. You can easily become depressed if you are isolated for a long period of time. Fortunately, we have the internet. I can’t tell you how many hours I’ve spent communicating with my family digitally – or playing online games with my friends during the pandemic. But it’s a lot.”
The coronavirus crisis has also been a challenge for Nikolaj’s manager Sidsel Holm Larsen.
“We have strong comradery in our little department, and communicate a lot with each other when developing and maintaining solutions. During the coronavirus shutdown, we had a lot of video meetings to maintain good team spirit. It wasn’t optimal, but it was okay for a while.”
She agrees with Nikolaj’s description of their workplace.
“Not to brag, but I think we have the best job in the world. PNP is a good example of the data-driven applications and tools we develop to help PostNord make better decisions – for the benefit of both our customers and ourselves. It is meaningful and it’s important for job satisfaction.”
On several occasions, the team of Sidsel, Nikolaj, Jesper Sangill and Jakub Kisel has received a round of applause during meetings with other units at PostNord – out of sheer enthusiasm for the solutions they have developed.
“Nikolaj should take a great deal of credit for that,” Sidsel says. “He is not only technically skilled, but also good at explaining what we do in a way that everyone can understand. I am in the process of hiring two new employees for our department. The demands are high, but if they have a profile similar to Nikolaj’s, I will be very happy.”
Sidsel Holm Larsen and Nikolaj Settnes
Position at PostNord: Sidsel is the manager and Nikolaj is a digital developer at Business Intelligence in Copenhagen, Denmark.
Closest colleagues: Jesper Sangill and