Welcome, AI!

ChatGPT would have started this article differently. But so he begins in 1990 with a nine-year-old boy who, when his parents are not at home, sneaks into their office and gets to grips with the computer: he pulls plugs on the huge square box, turns screws, opens the housing and swaps connections. When his parents come home, they are surprised by the smell of scorched cables... Patrick Jobst is curious, his enthusiasm for computers almost as old as he is. Computers and their technology exert a great fascination on him. It was logical that he decided to train as an IT clerk after business school. As a teenager who comes from Schönbrunn near Wunsiedel in the Fichtel Mountains and likes to go out in Hof, he often passes by the GEALAN building in Oberkotzau. "I don't even know what exactly it was, but the GEALAN look, the glass tower, somehow it seemed very attractive to me. After my civilian service, a position as an administrator was advertised – and I hired straight away." GEALAN not only offers Patrick Jobst the job he wants, but also development opportunities: He is completing an evening course in computer science business administration (VWA). "I wanted to improve my prospects." His career took the next leap in 2012: He took over team management and responsibility as Head of IT Operations. This includes the User Helpdesk, whose employees solve all kinds of computer problems; on the other hand, the administrator colleagues, who take care of networks and servers, among other things. "Our focus is on providing reliable and high-performance infrastructure – there are many great apps that all have one thing in common: they need a stable basis." Patrick Jobst and his team decide which computers and which mobile devices with which Internet connection should be networked and how, or better not networked for security reasons. The latter is important to Patrick Jobst, who is also an IT security officer.

Lucas Colling has taken a completely different path to IT. At 15, the Remscheid native likes to spend his afternoons with his school friends. Together they play shooting games. Graphics, speed and above all the competitive idea inspire Lucas Colling. To be faster, more skilled, more successful than the others – he often succeeds. In addition to gaming, there is also a real interest in computers: Setting up an operating system from scratch is no problem for him. Born in 1999, he is one of the digital natives who have been used to the Internet since childhood. Answers to computer questions are available online and Lucas Colling likes to get into computer topics. Nevertheless, IT as a career aspiration is far away for him. He enthusiastically plays theater at school and considers becoming an actor. The drama school in Essen would be his dream, but it is difficult to be accepted there. After graduating from high school, Lucas Colling travels to Australia for eight months – for a Work and Travel, not booked through an agency, but on his own. He quickly realizes: Getting accommodation, a job, a car is more complicated than he thought – he finds solutions and, as he says himself, learns a lot for life. Back home, the desire to act fades and Lucas Colling's interest in business awakens. He decided to study business administration at Hof University of Applied Sciences, chose to specialise in Digital Business, where he dealt with process optimisation and digitalisation, and came into contact with GEALAN during his bachelor's phase – for a practical semester. "I also got to know other companies from the region, but somehow everything was so vague with them. GEALAN had a concrete plan to use me in demand management, and also a clear idea for my bachelor's thesis, which really had practical benefits – that totally convinced me." In his bachelor's thesis, Lucas Colling is already dealing with the Microsoft Power Platform and Power Apps. As a working student in demand management, Lucas Colling uses the basis that Patrick Jobst spoke of and that he creates with his colleagues from IT Operations: He is looking for new, better applications for the various departments. "We often rely on a mix of make and buy solutions, i.e. we buy platform solutions with which we then build our own apps; we prefer low-code or nocode solutions, so we hardly have to program ourselves, but adapt the apps for our purposes." Patrick Jobst, the ex-hobbyist, and Lucas Colling, the ex-gamer, are now helping to decide how GEALAN brings artificial intelligence into everyday office life. Many office employees at GEALAN have been working with AI since 2024: they use chatbots such as ChatGPT for internet research, generate images with various AI assistants, and use DeepL as a translation tool. All this is now a matter of course. "Chatbots are so revolutionary and are so quickly and well accepted by all of us because their use is so simple and intuitive, almost human," says Lucas Colling. "I ask a question, I get an answer, like in a natural dialogue." AI tools are a great relief in everyday office life. So far, however, this has only applied to freely available tools. "This knot will soon burst," says Patrick Jobst. GEALAN will also make internal information usable for its employees via AI. "You have to imagine that we classify internal documents," explains Patrick Jobst. "Anything that contains personal data is of course removed from the access of an AI – it is extremely important that data protection guidelines are observed. All colleagues must be able to trust that their data is secure and will not be used to train AI." However, there is – and this is the crucial point – countless internal information at GEALAN that would be extremely helpful with access via AI – and this AI access will soon be available, with the help of Microsoft's Power Platform, which provides its own AI agent with the Copilot. "The premium version of the Copilot can then draw from all the internal data that we have expressly released," says Patrick Jobst. "If we open the barrier and allow the flow of data, completely new possibilities will arise."

One of them is the processing of the spoken word: "Minutes of conversations are cumbersome to create today, not to mention uniform quality," says Patrick Jobst. The right AI, on the other hand, will be able to deliver reliable protocols of identical quality in the future. But not only the protocol alone is helpful, the AI will also be able to summarize conversations, filter out important information, merge several conversations. If there have been four important rounds of talks on a topic, the AI will be able to quickly present the key results to an employee who wants to get information, without having to laboriously read all the minutes of the conversation. This makes it much easier to get up to date on a certain topic. The Copilot will also significantly improve internal research, because it will be able to access, summarize, filter and present data in different environments such as Excel, PowerPoint, Word, but also in Sharepoint, OneDrive or Webex, for example. Time-consuming searching in different programs and folders will then be a thing of the past. AI tools are revolutionary when it comes to generating, recognizing, excerpting and transforming text or images – especially the design and sketch layer will be much easier to create. Summarizing internal papers, using your own images and logos for new creations, but also translating documents so that they can be understood equally at all GEALAN locations – all this will no longer be cumbersome and time-consuming, but will work at the click of a mouse. AI will also play an important role in controlling: Many data analyses will be easier and faster.

Lucas Colling (left) and Patrick Jobst see great opportunities in artificial intelligence. In the future, they are convinced that every office employee will use AI like an assistant. The hunger of colleagues for AI is great. While Patrick IJTo Obspt eurnadti osnesin d Tiea Inmfr daestrr uktur schaffen, Lucas Colling identifies AI tools in demand management that make GEALAN processes even more efficient.

"We envision AI solutions, especially the Microsoft Copilot, as assistants that every office employee at GEALAN has at their side," says Lucas Colling. "Of course, he doesn't replace a single real employee, but he is an important help for everyone. AI will first and foremost take over stupid processes that always run the same way – the increase in speed and efficiency that this alone will result in will be very large." What artificial intelligence can do, however, goes far beyond that. "There are use cases without end," says Lucas Colling. "They are slumbering in the minds of our experts." A vivid example is the purchase requisition at GEALAN, which is always used when purchasing is to procure something specific. In the past, the purchase requisition was written on paper, and a form had to be filled out. In the course of digitization, the paper sheet became digital: Lotus Notes was used to create a form that was used to fill out the purchase requisition on the computer. With the end of the outdated Lotus Notes and the launch of the modern Microsoft Power Platform, we are now moving in the direction of AI at the same time: The individual items no longer have to be manually transferred to a form on the computer, but are directly recognized and inserted by an AI via text recognition as items, prices, control keys, etc. A connection to SAP is the next step – this makes the purchase requisition immediately bookable. "Purchase requisitions often have a lot of items. Nobody has to type them out anymore, it just happens automatically - that saves time, of course," says Lucas Colling.

Another example is safety data sheets. They play an important role in GEALAN's occupational safety because they contain precise instructions on how certain hazardous substances must be handled and tested in accordance with standards. There is a separate test procedure for each hazardous substance, which can be very complex. "This would be a good opportunity to use an AI agent," says Lucas Colling. "You could develop an AI agent specifically for this purpose, which is trained for every hazardous substance and then knows every process exactly. You tell him the hazardous substance and he spits out the specific information, the entire process that has to be followed."

It is important to Patrick Jobst and Lucas Colling that the ideas on where AI can be used efficiently come from the specialist departments. Leaving time-consuming, always the same processes to AI makes sense to everyone, but gradually more complicated processes will also get AI assistance. "The curiosity, I would even say the hunger of our colleagues for AI is great," says Patrick Jobst. "Anyone who has a full desk is grateful and open to new, fast and good solutions." "Our job is to bring together the expertise and the existing systems with the AI possibilities," adds Lucas Colling.

Patrick Jobst's field of work has changed dramatically in the 23 years he has been with GEALAN. "When I started, we were a mini-team of three people, everyone could do everything. Today there are twelve of us, we have created corporate structures within IT, we have driven forward networking with our parent company, which co-hosts many applications, and we have found suitable cloud solutions for us. One of our important tasks, even if it may not sound so sexy, is standardization and consolidation – we want to have solutions that are as similar as possible at all GEALAN locations and are well networked. I also see this as a form of professionalisation." GEALAN-IT has specialised and now has more tasks than ever before. "My job is exciting every day. The environment is constantly changing."

A few years ago, no one could imagine working with artificial intelligence. Today it is the other way around. "No one can know what will happen in five or ten years," says Lucas Colling. "But the fact that AI makes us all unemployed: that certainly doesn't. Natural intelligence is needed especially for complex decisions that have to take many factors into account. We are still a long way from the superintelligence that can think and decide like us." If ChatGPT had written this article, this quote would probably not be here.