“It is important to speak the same language”

Koen Peters, alumnus and external PhD researcher at Tilburg University, is a remarkable data scientist. However, his work on his PhD research is taking a back seat at the moment because he is also working full-time as a consultant for the World Food Programme in Rome.

His passion is applying math in unusual places. The World Food Programme provided a big challenge: the analytical nature of math is hard to combine with humanitarian aid. However, a tailor-made mathematical model was necessary to make the Food Programme better and more efficient. That was only possible by winning hearts.

Speaking the same language

Koen entered into a dialog with various parties within the WFP, each with their own Excel data files, to map the entire food supply chain, from the composition of the food package to purchase, and from shipping to distribution. A mammoth task, which he performed while working as an intern.

“It is important to speak the same language. The more concrete an idea is, the better it is to understand,” he says. In his opinion, that was the key to the success of his model for the WFP. He had already experienced such a consultation process during the Outreaching Program for talented students at Tilburg University. And it worked. “It takes some persuasion, because everyone wants something different. You need to be able to empathize with people.”

Highlights

One of the highlights was testing his prototype for the model’s final version at the end of his Master’s thesis project. For the ultimate test, he was sent to Syria with the prototype to talk to the WFP supply chain managers. It was exciting: how do you prove that your model actually works? But the meeting yielded concrete suggestions for improvement that could help save the WFP millions of US dollars: 20 million in Syria and later 1 million a month in Iraq. “We got nice feedback, for instance, that the food package was a better match with the normal diet of the people in Iraq.”

Koen sees opportunities to further develop the model for different situations, such as arranging government-sponsored school lunches in India. He is currently working as a consultant at the WFP in Rome on making the model accessible for staff who are non-experts in data science. And on his PhD, of course. On weekends.