
When you talk to K2 Head of Development Tomáš Szkandera, his excitement about how far the use of artificial intelligence in K2 has come is evident in every sentence. “This is a genuine revolution in enterprise systems,” he says enthusiastically. Artificial intelligence has become one of the fundamental building blocks for the future development of K2.
One of the most significant developments is K2 Lisa, the virtual assistant that was already available in the previous version of K2. At that stage, however, it worked with a limited range of data. “We taught Lisa the entire K2 documentation as well as the content of our recorded webinars. This made it easier for users to find answers without searching through help documentation or calling our hotline,” explains Tomáš Szkandera.
In the new version, Lisa can work with data from across the entire system. This enables her to perform analyses using virtually any data available in K2. “It works in the way users have become accustomed to with the most popular and widely used language models. You can have a natural conversation with Lisa and ask her anything you want to know about the system. Within seconds, she can answer a manager’s question such as: Which store generated the highest revenue or margin yesterday? At the same time, she automatically suggests follow-up questions and more advanced business analyses,” adds Szkandera, who has led the K2 Group’s development team in recent years.
“Even today, we provide owners and decision-makers with an extraordinary amount of data, giving them a comprehensive view of inventory, production, orders, financial indicators and accounting operations. We also work with standard dashboards created in Power BI. In the new version, however, everyone can access information in an entirely new way. All you have to do is type – or even simply ask by voice – what you want to know. Within seconds, you receive clearly presented information, complete with useful context and suggestions for potential next steps. It is like having Power BI available within seconds, empowering managers to become much better decision-makers. You no longer need an in-depth understanding of the system itself. You simply need to ask the right question and then use the data to make your next decision,” says Petr Schaffartzik, Chairman of the Board of the K2 Group.
Unlike publicly available language models, an enterprise system must, of course, comply with the strictest security requirements. “No sensitive data leaves the system when artificial intelligence is used. We have developed everything ourselves, and all analyses are performed exclusively using internal company data already processed by K2 software. Access is governed by precisely defined user roles and permissions, just as it is today. We see security as requirement number zero – it is something we take into account automatically. If something is not fully secure, it simply does not move on to the next stage of development,” explains Tomáš Szkandera.