Dear future colleague, I never planned to become a data scientist

A letter from your future colleague, Mikko
A letter from your future colleague, Mikko
May 17, 2018

Read time 6 min

Dear future colleague,

Even though we are living in the midst of data hype, the question I often hear is: “What is it that you really do?”.

Now that I am sitting here at our Helsinki headquarters, unwinding after a stretch of intense preparation and a successful demo, I’ll try to shed some light on what we as data scientists do, what our projects are like, and give you a glimpse of what to expect when you join us.

There’s no one path to follow to become a data scientist

I never really planned to become a data scientist. Originally I wanted to be an engineer but became a neuroscience researcher instead. Sometime after getting my PhD, I realized that I was more interested in setting up the experiments and playing with the data than the actual experimental work. I had a background in mathematics and some coding skills, so I eventually found myself immersed in statistics, models, coding, and thinking about how I might use these things. Then I got a new job as a data scientist, first somewhere else and then at Reaktor.

My path to becoming a data scientist might not be a typical one (if there even is such a thing). The backgrounds of Reaktor data scientists are varied, and that is a big benefit. We work with real-world problems, with real-world data, and actual business problems of our clients. The knowledge we have of different fields helps in understanding the contexts and the possibilities in our projects.

At the moment, I could not be happier. I get to use everything I’ve learned during the past twenty-odd years. I get to draw from my experiences in leading a team of highly skilled and motivated professionals, from all the scientific ways of thinking I’ve accumulated, from the domain knowledge in fields I’ve worked in and from what I’ve learned about interaction skills, professionally and privately, and put it all to use. I get to dig deep into the depths of new technologies and the models we build and, at the same time, ponder over some big questions: what comes next, where are we heading, where can we take this?

A good data scientist is like the drummer of a punk rock band

Now, what is it that we do then? The landscape of what can be seen as data science is vast, and so is our project portfolio. It includes everything from a single data scientist hacking away and finding insights from limited amounts of data to full-blown teams of data scientists building complicated cloud infrastructures and handling tens of millions of events daily as source data. From giving lectures and writing pieces to a variety of forums to helping entire organizations transform to fully utilize the possibilities of data science and new ways of working.

As technical solutions go, we do everything: straightforward regression modeling, computer vision, complicated natural language processing, and so on. The projects span a range of industries, from retail to healthcare and manufacturing.

As you see, technology-wise, we do a bit of everything. But that’s really not what it is all about. Sure, you need to know your way around the tech and the numbers, but what we really do is solve problems and help our clients use data to create value. And this is important. It’s the mindset behind our thinking when we approach a project. It’s not about the coolest models or the latest technology, it’s about finding the real problems, identifying the ones that are worth solving and figuring out the most suitable and fastest ways to solve them. That entails a whole lot more than just programming models. We also spend a lot of time educating people on the possibilities of data science, on thinking what the real problems are and how could we tackle them.

A good data scientist is like the drummer for a punk rock band. There’s no place for a solo when all that is needed is a solid D-beat (you have to know how to play fast though). The state-of-the-art 50-layer classifier you spent five months building and training might not be the right solution if logistic regression would have solved the problem in five days. On the other hand, if the project calls for it, you can pull off the latest reinforcement learning algorithm that has never been built before – and do all this while playing together with the band.

Make learning a priority and a habit

Starting at Reaktor will probably be a whole new experience for you. You’ll be greeted with a very welcoming atmosphere, with loads of people wanting to talk to you and to help you out. You get to meet a lot of talented experts who want to get to know you and how you think. Your opinion is valued from day one, so don’t be afraid to say it. We want to learn from you just as much as you want to learn from everyone here. This thirst and yearning to learn new things is something that we share at Reaktor, so embrace it! I know it might feel like a lot at first, but before you know it, it’ll be the only way you know how to be.

Maybe you have a specialty: can make computers see in complete darkness, or have your laptop understand speech in a matter of hours, or are so Bayesian you can’t finish a sentence without referring to the prior. Great! Someone will try to find you a project where you can use your magic. That being said, we value it when you step out of your comfort zone. Whether it’s hopping into a project in a new-to-you field, training people in the intricacies of AI, giving speeches or writing blog posts, everyone will be happy to help you out to expand your horizons.

Then after a while, you’ll be thrown into – or not really thrown into, more like you’ll be dying to jump into – the world of sometimes messy and missing data, expectations, problems and solutions, and all the other things found in our numerous projects. It might even happen that you are the sole representative of your craft present.

Don’t be afraid. First: you can do it. You’re smart, you’re resourceful, it’ll go well. Second: you’re a Reaktorian. That means that you have the whole of our data science community of practice, as well as our wonderful developers, designers and everyone else backing you up.

The one thing you can safely leave at home when coming to Reaktor is your ego. Trust me, the only time you’ll be the smartest person in the room here is when you lock yourself in the bathroom, alone. This company and community feed off of collaboration and openness, lone rangers and rock-star attitudes don’t go far. Your best aids will be your two ears. There are a lot of wise words said here, so make a habit of listening and absorbing.

Now I have to run, but I’d like to see you soon over a cup of coffee at our office or working together to wrap our heads around a challenge together. In the meantime, if you’d like to hear more about what it’s like to work as a data scientist at Reaktor, you can reach me at


P.S. You’ll find our open positions here.

Never miss a post