An interview with Kristian Ovaska — Valuemotive’s super data scientist
Valuemotive in brief:
Valuemotive offers expertise in the fields of Data Science, Design and Software Development. We work with client companies, helping them build a deeper understanding of their customers and discover insights to explore new business development opportunities.
What is your name and can you tell me a little bit about yourself?
I’m Kristian Ovaska. I am 39 years old. I am a data scientist and a software developer.
Kristian at the Valuemotive office
What do you do in your free time — any hobbies?
Well, I do play a lot of computer games, which is a very common hobby that a lot of our colleagues in our office share with each other. Out of all the computer games I’ve played, I like strategy games the most — open-world, exploration sort of games. I also enjoy playing some farming games on my mobile phone. Sometimes, I do sports too. As well as that, I also like photography — I often walk into nature and have my camera with me — if I see something interesting, I would take a photo of it. I guess those are the main hobbies. I used to read a lot when I was young but I haven’t had a chance to read these days anymore.
Matias and Kristian trying out Tier Scooter for the first time during our lunch break
What does a ‘‘data scientist’’ actually mean?
Another word for being a data scientist is a “data analyst”. I usually work with clients’ data, on which I perform data analysis and machine learning. So it is definitely not like you are doing science, more like running an analysis with the data. In short, my main responsibility is to extract value from the data that our customers provide us with.
What are the differences between a data scientist and a software developer?
In my opinion, there is a noticeable difference between a data scientist and a software developer. Data scientists and software developers have related skill sets. The difference is that data scientists tend to focus more on finding value from data, whereas software engineers pay more attention to building quality software. Nevertheless, data science requires building custom software — all data scientists also need software engineering skills.
More specifically, a data scientist is a professional analytical data expert who possesses the technical skills to solve complex problems and figure out ways to explore what problems need to be tackled. They are responsible for data collection and data analysis in order to identify various approaches to improve business operations. Furthermore, data scientists are able to explain what is taking place by processing recorded historical data. They can also use various advanced machine learning algorithms, which aim to recognise the occurrence of an event in the future. This will in turn help in decision making and predictions. For this process, a data scientist has to look into data from various angles.
On the contrary, a software engineer is an individual who possesses a knowledge and applies the disciplined, structured principles of software engineer to all the levels — design, development, testing, maintenance, and evaluation of the software that will avoid the low quality of the software product.
Can you briefly describe your daily tasks? How would you describe your normal working day?
My daily tasks vary quite a bit and they often depend on the client project I am working on. For instance, I remember there was a project that I was assigned to a few years ago. Without going into too much technical detail, basically they had this website where they sold various products and there were lots of different types of products. One of my responsibilities was predicting the market price for those products.
There are different factors/metrics determining the market price of a certain product — manufactured prices, product conditions, colours, etc. In data science and machine learning, we learn these factors and their behaviour automatically from existing data using algorithms. You have to manually insert a lot of example data into the algorithm from the database — about 10000 product details that other sellers have set the price — so the algorithm will learn that, for instance, this brand Y will have higher prices than other brands. And if red-coloured products are priced higher than those black-coloured ones and algorithms will learn these rules too.
Daily analysis process is something that I have had a lot of experience working with. Generally speaking, we use standard algorithms in data science but you have to apply them to a specific problem. It is mainly because algorithms are quite general and in principle one can pretty much apply them to almost any kind of data set. However, what I do is that I have to tell the algorithms how it can find out if the product is red or black — what is the structure of the database — so a lot of adaptation and configuration to be done. So basically, I am writing custom code and it’s software development in a nutshell. That’s why I like this dual role: data scientist and software developer. Even when I’m doing this kind of so-called pure data science, I am developing the code at the same time. It is because not only do I have to code the algorithm itself and things, but also how I insert data to that algorithm and how to interpret the data when I got the results.
Towards the end of the project, I usually write a report for the customers based on the results. I think it would be valuable for our customers to find out, on average, if red products are more expensive than their blue counterparts. I would then convert the results to PDF documents, Excel files or in some other cases, I can build small-scale websites for our customers to browse the results.
How long have you been working at Valuemotive? How do you feel about working at Valuemotive? And do you like your job?
I do very much enjoy my job, otherwise I would not be here in the first place. I have been working at Valuemotive since 2014. The thing I like the most about working at Valuemotive is the fact it has a very dynamic working environment. I am the type of person that would certainly get uninspired if I am stuck in the same project for an extended period of time. So at Valuemotive, my projects and clients will change after a while — so there is a lot of variety. When I work with different customers, they usually have different types of technologies and I have my technology toolboxes. However, if the customers have their own infrastructure, then I have to figure out ways to integrate my own technology into their existing systems.
Working with different types of clients provides me with an opportunity to learn more about the technology available in their company. So for me personally, learning aspect is the most important element. Even though you are an experienced data scientist, when you go to a new client, there is always going to be something new that you might not know before. It is mostly because today’s technology is always going to be slightly different from yesterday’s technology. It is not the same as you only have to build a software once because the next customer might have a completely different set of data, technology and, often times, business needs.
I once built for a customer a software to predict the market price of a product and that software is a one-time use. For other customers, I also do software development for their projects, but so not as much data science I’d say — just basically building more software, which adds to the whole dynamic working environment aspect.
One typical weekly meeting at Valuemotive
Challenging aspects that come with my job
My job is all about IT consultancy and working with different customers/clients, which I enjoy. However, it can also mean that I often work to solve the customers’ problems and improve their products. Even though I love the dynamic aspect of my job, I do not have any ownership over the products that I created. It would be a different situation if I were to be working in a company that does machine learning, I would feel a stronger sense of ownership because I would have the mentality of “it is our thing and we are improving our thing”.
However, the fact of the matter remains when I am working on customer X and I will be improving product X. Some of my colleagues are working for customer Y and therefore improving product Y. I believe it is a trade-off type of thing. With that said, I’d still prefer this kind of dynamic way of working.
Why did come and work for Valuemotive? How was the whole process like?
I came here to work because I know Marko Laakso — one of Valuemotive’s most experienced data scientists and our partner. Also, I have known Pekka — our CEO since 2011 and we did a lot of collaboration back then. I basically already knew quite a few people before I started working at Valuemotive.
(Elias), Pekka (orange shirt) and Marko (gray shirt) after a cooper test
What are the advantages one will have when working at Valuemotive?
I actually quite like the fact that Valuemotive is rather small-sized — we are a company of 25 people and we have a flat organisational structure. Admittedly, there are advantages when one works in bigger IT consultancies. However, it is just so easy for one to be lost in a big mass of people [when working in for a big consultancy]. Speaking from personal experience, in a smaller company like Valuemotive, if you have some ideas about how to develop business, sales, or you are not happy with something concerning, for instance, your well-being or your project, you can just say it. Whereas, if you work in a company that has 500 or 1000 people, it might be hard for you to have your voice heard. More importantly, I also believe that it would be much easier for one to have an impact on a small company where there is no hierarchy whatsoever.
Friday afterwork beers with some of the Valuemotive gang
Conflict resolution tactics at Valuemotive
I don’t think there have been many conflicts in Valuemotive during the time I’ve been here. It is mainly because the majority of Valuemotive’s consultants are working with our customers. I’d say I have spent roughly 80% of my time at clients offices. So if there are any conflicts happening, it is often times related to the projects and the customers, instead of any internal conflicts happening at Valuemotive.
For instance, if there is a certain degree of tension between team members and the conflict is not resolved, I usually go to Pekka and talk to him about it — it usually helps. It is usually the case, particularly if the conflict happens in our team that was formed to work on a certain client project.
What would you say are some of the biggest challenges you face here?
From my personal perspective, the most obvious challenge is finding ways to create a sense of work culture in Valuemotive. For instance, when the majority of us work in customers’ premises, it becomes increasingly challenging for Valuemotive to build and maintain a unified company culture if you only have one-fourth of our people are working from our office and the remaining lot stay at the customers’. However, it goes without saying that we will always get the project done better when we are at the customer’s premises.
One solution that I can think of that could potentially change our current situation is when we can sell more of our own products, meaning Varis or Pyy or larger-scale IT services, we can then have more Valuemotive employees working on the same project. I believe that’s the challenge for people in our sales department.
Why should customers use IT services from Valuemotive, in your opinion?
Generally speaking, we are relatively strict when it comes to recruiting our employees. So we usually have pretty skilled employees, they are all motivated and capable of teamwork. As an active participant in our hiring process, I’d say Valuemotive’s employees are either competent data scientists/software developers/software architects with lots of experience under their belt or exceptionally talented junior developers who have shown lots of potential in their technical knowledge and business commercial awareness.
So for me it is safe to say that the quality of personnel is rather high in our company. We also have the traditional software development and as well as that, there is data science, data analytics, machine learning and service design aspect, so you could say that we can offer our customers a full package of services.
What are the most unique elements about Valuemotive regarding our product offerings?
What is most unique about us is Natural Language Processing capability (Varis and Pyy), which is the field that we have the most expertise in and feel most passionate about. For instance, we can say with absolute confidence that we offer the most modern and state-of-the-art Natural Language Processing (NLP) capability to all of our customers. However, a customer may need something else built, like a mobile application, that does not concern natural language processing. It is pretty much why it’s crucial for us to [be able to] sell this particular type of software development.
Is there anything else that you would like to share about Valuemotive (any experiences, any memorable stories, etc.)?
During my time, I have witnessed rapid changes at Valuemotive and for the most part, those changes are for the better. I believe that Valuemotive’s working culture has positively improved. I have started hanging out more and more with my colleagues at Valuemotive recently. When I first started working here, I still remember spending the majority of my time at our customers’ premises and I hardly had any chance to meet my other colleagues until a Christmas party.
In addition to that, the business aspect has definitely improved, in the sense that we used to be a traditional software development consultancy, but now we have our own unique analytics product offerings, e.g. Varis and Pyy*. So that’s a very good sign to tell the world that we have made decent progress over the years. So all in all, I feel quite optimistic and hopeful about the future of Valuemotive.
Finally, what would you say to Valuemotive’s potential employees?
Perhaps the dynamic way of working is the most important thing — you have to always be willing to learn, to get used to the changing nature of different types of projects and customers. So if you are the type of person who likes to stay in the same kind of project, the same workplace and using the same sort of technology for 4 or 5 years, then Valuemotive might not be the most optimal place for you.
On the other hand, if you are the type of person who likes to learn about new things, meet and make friends with new people — Valuemotive will soon become your second home. Take myself for instance. Since I have been in a lot of different projects and countless teams, I have had a chance to meet and network with a huge amount of people — and because of that, I have earned lots of new project leads, contacts, as well as long-lasting friendships.
Valuemotive’s gang and beautiful Finnish sunset at Mökki
Varis and Pyy*: For those of you who are more curious about Varis, please head over to https://www.varis.ai for more information. For Pyy, stay tuned, guys!