Shaping AI instead of reacting to it
KI-Hackathon

Shaping AI instead of reacting to it

AI is evolving not just day by day, but almost hour by hour. At this point, no one is debating whether to use AI anymore—only how to use it—and that goes for us at AEB as well. So what happens when a group of people takes a day to not just react to these changes, but to actively shape them?

“The future doesn’t just happen. We create it,” writes Prof. Hannah Fry in her book “Hello World: Being Human in the Age of AI”. The AEB Customer Solutions division has taken up this idea and made it the theme of its first AI hackathon. The goal: to promote the use of AI at AEB while steering it toward responsible and value-adding applications.

Shaping tomorrow's work day today

“Just give it a try” is considered a given in the software industry, but according to hackathon co-organizer Hannes Eckelt, it’s more important than ever: “Even in the past, before AI, you always had to be ready to set aside what you knew and embrace new technologies. The same applies to AI now, only to an even greater extent, because it’s completely transforming the way we work.”
AI is changing—and AEB is changing with it. That’s why the hackathon focused on the question of how and where we can best apply AI, from the initial idea through to the implementation of our projects.

“The world of AI is too vast for any single person to grasp, and it’s growing by the minute. That’s why there are people who identify the truly essential aspects and make them available to project developers.”

– Tobias Weiß, co-organizer and software developer

At the hackathon, therefore, a general openness was more important than technical expertise in a specific field: on an individual level, openness to new knowledge, and on a collective level, openness to new approaches — whether by bringing your own idea or joining an existing team. This resulted in a total of 15 teams for the day, covering a wide range of topics: from the further development of an AI process for summarizing sales calls and follow-up processes to the improvement of existing AI assistants.

Collaborative development, individual learning outcomes

The participants deliberately included people with a variety of roles and levels of AI expertise. This was no coincidence: the team composition followed a clear idea.

“The goal was to give people the opportunity to experiment with AI tools and familiarize themselves with this new technology by providing them with a starting point where they could try out specific things. At the same time, the hackathon gave experienced users the space to put their knowledge to good use and create something they believed others could benefit from.”

– Franz Gaber, co-organizer and project manager

The participants also considered this approach a success: “I liked the diverse team composition because it allowed me to work with people who had more prior knowledge than I did,” says Filippa Kouthoofd-Lans, a project manager in the Customer Solutions division. “For example, I learned how to craft prompts and leverage the agent to improve how my team and I work day to day”

In addition to diverse areas of responsibility and levels of AI expertise, the hackathon also brought together multiple locations. Project managers and software developers from Stuttgart and Malmö organized the event together, from planning through to implementation.


The use of AI needs room for experimentation

Between project work and deadlines, there’s often little time to optimize processes. In the long run, however, automation can save a lot of time. That’s why the Customer Solutions division consciously invested a full day in its own development.

“I think it’s great that we had a day dedicated to workflow optimization. That way, you don’t have to deal with it amid the daily grind when you don’t necessarily have the mental bandwidth or resources. Plus, at the hackathon, you can ask for help from people who’ve already delved deeper into the topic.”

Marco de Jesus António, participant and software developer

The teams' presentation showed just how much can be accomplished in a single day when given the space to do so. In the long term, the event laid the groundwork for a better flow of information throughout the entire process — from customer requirements to implementation—and, in a relatively short time, empowered many people to use AI more confidently in their day-to-day work.