HackAIthon 2026 – How AI is Transforming Daily Engineering Work

During HackAIthon, each team was assigned a topic to work on throughout the event. Our tasks required combining technical expertise with creative use of AI – from planning and specification to prototyping applications that could genuinely support daily engineering work. Each of us had a clearly defined role, and our collaboration was based on exchanging ideas and continuous feedback. In this short time, we learned how quickly AI can turn our concepts into working solutions and how valuable it is to propose new ideas and iteratively implement further improvements.

Together with the team, I focused on the topic of SPEC-DRIVEN development – an approach in which a precise and as complete as possible specification is created before even a single line of code is written, becoming the main point of reference for implementation, testing, and documentation.

In our case, the ban on manually writing code was absolute – everything had to be created with the help of AI.

After two days of intensive work, my team and I managed to deliver a working first iteration of the application, run tests, fix minor issues, and prepare the final presentation along with the PITCH – all with AI support. Although our solution placed 9th in the participants’ vote, the satisfaction from the results was far greater than the ranking itself. HackAIthon showed me that AI not only streamlines work but also allows experimentation at a pace that once seemed impossible.

The most valuable part of the event for me was meeting people who share a passion for new technologies. The inspiring ideas from other teams and the energy that filled those two days made participating in HackAIthon a truly exciting and inspiring adventure.

I, together with my team, worked on topic: “AI Developer/Migrator.” Our goal was to create an application that supports repository modernization and migration – including upgrading Spring Boot, Java, and Gradle – as well as assisting migrations from Merlin to Mercury, our internal PaaS platform at RBI.

During HackAIthon, we had the opportunity not only to work on our own project but also to observe the solutions developed by other teams – and a few of them truly impressed us. One team created a deployment support system using specialized AI agents – one classified risks, while the other, in case of a failed deployment, performed a full analysis and provided a diagnosis. Another team used AI to automatically generate documentation in Confluence based on the code repository. The takeaway is clear – what seems futuristic today could become the standard tomorrow.

We were particularly interested in projects focused on practical, day-to-day support through AI. For example, the winning team presented RBI Kargo Agents for Mercury users, agents that assist with promotion decisions, risk analysis, and diagnostics in tools like ArgoCD. Second place went to a concept for automatically generating documentation in Confluence based on changes in the codebase, and third place – RBI Cowork, an environment where agents could perform end-to-end tasks without manually entering steps in the terminal. Congratulations to all the participants!

Meeting people who share a passion for new technologies, seeing the inspiring ideas from other teams, and feeling the energy that filled the event made those two days a truly exciting and inspiring adventure. We were surprised by how important clear goals and well-defined inputs/outputs are for AI – this allows models to be treated not just as advisors, but as a real part of the pipeline. There was a strong sense of “we build and show,” not just discuss. HackAIthon was also a fantastic opportunity to meet people working in AI and to exchange experiences.

We realized that artificial intelligence can genuinely support everyday engineering work if introduced within a well-thought-out process. Properly designed LLM models can become an integral part of the workflow, supporting migrations, testing, or documentation automation, rather than just acting as advisors. The event also highlighted the importance of collaboration, rapid prototyping, and knowledge sharing – from planning, through implementation, to presenting results. HackAIthon is a true lesson for anyone who wants to use AI in practical projects and experiment at a pace that, just a few years ago, would have seemed impossible.