Technology

Latvia’s AI Revolution

For Boardroom, Girts Smelters from Latvia-based Exponential Technologies (xT) explains how AI helps find the right recipes for industrial innovation and why connecting local knowledge with international meetings makes a lot of sense.

Can you explain to amateurs like us what kind of tech products you have been developing? 

xT is a growth-stage start-up, founded in mid-2019. Our core business is to provide an AI platform and research management system that allows industrial chemists, biologists and engineers to find and optimize process/machine parameters, material composition and many other parameter types. It was founded by Pavel Cacivkin, Matthias Kaiser and myself, and we have a strong and growing team across offices located both in Riga and Rezekne.

So that you can better understand what we do, I can draw an analogy from our daily life.

Let’s assume that you want to open a fast food pancake restaurant, and you need to find the perfect pancake recipe – one that is extremely tasty, easy to mix, quick to cook, and with ingredients that are not too expensive. First, you’d have to find an optimal ratio of baking powder, salt, sugar etc. Then, the right sequence of how you mix them, following with the discovery of right process parameters – frying temperature and duration. And you would have to do all of these steps simultaneously, while spending as little money as possible. Traditionally, any seasoned chef would start by trial and error, changing one or two parameters at a time – a little more sugar, a little higher temperature and so on. That’s why it takes decades for people to reach true greatness.

Our AI platform can substantially shorten this time and save a lot of expensive resources, as it will help to discover the most optimal solution rapidly and let the restaurant owner move on with the business operations and new additions to the menu. Our AI solution starts learning from scratch by generating suggestions that you have to actually try out, and expects your feedback about the taste, performance and costs. This is done for several iterations, and, finally, the AI arrives at the desired result. That is called multi-objective optimization.

We suppose the pancake example is a simplification…

Correct, but once it comes to other industrial domains or medicine (like vaccines against COVID-19) the ingredients and processes become very complex. With the help of our AI solution xT_smart_DoE you can speed up the discovery and production scaling for vaccine and antibody test kit manufacturing. This would shorten the go-to-market time for COVID-19 vaccines by several months, and thus save tens of thousands of lives. Hence, we are currently negotiating possible collaboration with some of the global vaccine manufacturers.

Girts Smelters, courtesy of LIAA

Presently, we focus on biotechnology, chemicals and advanced manufacturing. Applications are possible e.g. in pharmaceutical and chemical production, but also for the production of sustainable food and beverages (like, cultured meat), bio cosmetics, smart materials, novel composites and alloys and so on. The core engine of the platform are AI algorithms which we have developed in-house. 

However, the service is tailored for each domain specifically, offering a comprehensive research management system for teams of industrial researchers and engineers to collaboratively innovate. Our business model is SaaS (software as a service) on the cloud or can be provided as an on-premise solution — as industrial production often involves very sensitive data and needs to be kept offline. We believe that the opportunity to innovate should be as accessible as possible, therefore our aim is to build products in a user centric manner with an ease of use in mind.

How do you find clients or potential clients? 

We are very active in attending conferences, expos, all available platforms and we use any opportunity for networking. xT is part of the Latvian delegation group at Node by Slush, the recently launched virtual start-up community. We will be there from October – December this year, you can visit us here.

In addition, we actively take part in hackathons. xT, together with a team of international scientists recently won the Health+Life (research category of the EUvsVirus hackathon) organized by the European Commission. In total, the competition attracted over 20 000 participants and over 2,100 projects were submitted. We had a lot of good press coverage in local media here in Latvia! Participation in hackathons has also served as a great networking tool and a possibility to gain international exposure and recognition in the times of global lockdown.

Where are your clients based? 

We find them in all big economies — Germany, France, UK and Spain, but also in the USA, Middle East and quite recently we have started to explore the Indian market. Latvia, predominantly with the help of LIAA endeavors, has built up a strong position in the regional deep tech scene through a very systematic approach and invests a lot in early-stage ventures. That creates favorable conditions for startups to emerge and grow here. However, the scale up gets challenging, as it requires proximity to large industrial players in order to run pilots and approbate the technology.

In order to promote meeting and congress destinations, one usually present knowledge and innovation in the destinations to create a new type of context and relevance. Does this make sense to you?

It does make a lot of sense! For example I attended Smart City Expo World Congress 2019 in Barcelona and they offered all kinds of relevant opportunities to meet, network and visit organizations related to our interests. I found this approach very inspiring and in fact it helps to connect much stronger to a place, adds value to the congress participants and visitors and expands the networking opportunities. 

Hit enter to search or ESC to close