Meet Akseli Kelloniemi: The Green Sage of AI

"I’ve been a software developer for the most part of my career and come to realize that one of my inner motivations rises from creating value, not from the actual programming itself. At this point of my career, software development flavored with AI/ML elements seems to be the next logical step and Veracell fits that requirement perfectly. A startup that is focused on AI while still having strong emphasis on traditional, cloud-based software development is exactly what I’m looking for. For a company that is relatively young, a group of very talented people sure made an impression on me as well."

As one of the main drivers at Veracell is to do things that are useful and valuable, thus Akseli was a perfect match for our company.

"I’ve had multiple identities over the years and in addition to a coder, I currently identify myself as a father and an overall nerd. I enjoy being by myself and working remotely has given me an opportunity to be able to focus 100 % on work without much disturbances. However, I do think that having a team around you is essential for self-improvement and quality of work.

Currently our toddler usually keeps me busy during my free time but if I have the chance, I like to tinker with the data I've been collecting about myself. In my latest project I've set up a system which records me while I sleep and then feeds the images to a neural net that is trained to identify the position that I’m sleeping in. Combining this data to other metrics has already given me a lot of insight.

With us, everyone gets to decide how and when they work, only results count. For those who like the company, our cozy office is always open right next to Keskustori at Tampere.

Sustainability is also close to my heart and for example, I’ve set up different sensors and energy meters around the house to model our house’s energy consumption, in order to select the most optimal heating source (air pump/district heating). When I have the time, I’ll try to utilize solar panels in a more effective way by automating certain appliances. I've also been in a leading role developing AI solution for finding optimizations for buildings' heating energy consumption. It is estimated to save customers yearly over 2 millions euros in heating expenses and nearly 5 million kilograms of CO2 emissions.

I never really dreamed of being a developer and I actually ended up in the university because my friend thought I’d do well there. Even though I spent most of my youth with computers, I got to know programming only through my signal-processing studies and then something really clicked. I felt like I was good at it and I immediately recognized the immense power that being able to program gives: ability to automate and develop new products with basically zero capital only limited by your own imagination is a really powerful feeling."

As Veracell focuses on cloud, data engineering and AI, Akseli brings a lot of experience on board. We are also continuously looking for data engineers and scientists to join Akseli in our team.

"I have a strong background in developing scalable, cloud native SaaS software with microservices architecture. Also in the past I've had to deal with massive amounts of time series data, experience that might come handy especially when developing software that's built around the data that IoT devices generate. In addition, I've built dozens of different integrations which provide a skillset that can be useful for data engineering tasks.

What comes to machine learning and AI, I’ve had some success already but there is a lot to learn. The field is huge and I think businesses are sitting on massive data sets with value yet to be discovered. In order to get anything meaningful out of the data, this is where data engineering comes into play. Although I do have some experience on it, there is a lot to learn about the latest common practices and technologies.

Currently I’m working on a pilot project as a data scientist. We’re trying to forecast demand for products using the available sales history but in addition, we have an interesting dataset which might have predictive value."