Aili Asikainen

“Data Scientist” has always had multiple meanings, it has a slightly different definition depending on who you ask. I think that inconsistency describes my work quite well.

…that the technology isn’t the point. The point is what you want to achieve and AI is a tool along the way, not the destination. “Think” is doing a lot of work here. You still need to understand what you’re doing and why, even when AI is handling the execution.

I’m drawn to problems that involve some level of optimization. Solving those problems requires deep understanding of the business case, clear communication about tradeoffs, and building a technical solution that can evaluate different options in a reliable way even when there’s uncertainty.

People I’ve worked with might describe my approach as: practical. I like to stay close to what matters and try to find a solution that works now, not in two years when the data is clean or the model is perfect. This means understanding the business problem and finding the parts that matter the most.

Like many data scientists, early in my career I thought that a good model spoke for itself. After a couple of prediction projects, I saw that people, myself included, didn’t really know what to do with the prediction when it’s presented. Working on two different optimization projects close to the business changed that. I realized that just by linking the output directly to the decision it was easier to work with. The value was never in the model. It was in the decision it served and supporting the decision makes a bigger difference than a perfect model.