Introduction
A hackathon is a time-bound, collaborative event where people gather to rapidly develop solutions to a defined problem (Briscoe & Mulligan, 2014). Originally a mostly tech-oriented format, hackathons have expanded beyond the technology sector and are now widely applied across many fields, including education, healthcare, and social innovation. What was once a space for developers is now used to bring together people from different disciplines to collectively tackle a common problem, making hackathons a uniquely cross-functional environment for innovation.
Earlier this year, I participated in a hackathon organized by Gustave Eiffel University in Paris. Our team had seven members, each from a different background, together representing four countries across Europe. Such opportunities bring together diverse perspectives and ways of thinking, which naturally generate creativity and push ideas in directions a more homogeneous group might not reach.
The theme of the hackathon was to solve transport challenges in low-density areas in Europe, a real, complex, and underserved problem. The goal was to develop an innovative solution, create a business model, and pitch it to a panel of judges. A working prototype is a powerful way to do this, not only as a presentation tool, but because the process of building one forces teams to test assumptions, define features, and gather feedback in real time. This is where we decided to use AI. In an environment where speed matters and ideas need to be brought to life quickly, AI became a practical tool for turning concepts into something visible, testable, and improvable.
From Problem to Idea: Narrowing the Focus
When our team first came together, we had a broad problem and seven people with very different professional languages, disciplinary backgrounds, and cultural perspectives. Early discussions were energetic but often unfocused. We were generating ideas faster than we could evaluate them.
The first step toward clarity was narrowing our target group. The transport problem in low-density areas affects many different people in different ways, and trying to design for everyone typically means designing well for no one. This is a core principle of service design, meaningful solutions begin with a specific and well-defined user (Stickdorn & Schneider, 2011).
Through discussion and with guidance from our coaches, we narrowed our focus to older people. This decision was driven by both need and relevance. Europe’s rural and low-density areas are ageing rapidly, and elderly people in these areas often face a combination of limited transport options, reduced mobility, and lower levels of digital literacy. Accessibility emerged as the central theme of our work. Designing for accessibility in this context meant considering physical, cognitive, and social barriers rather than focusing solely on technology.
From there, our coaches encouraged us to narrow the geographical scope further. We focused on Creuse, a rural department in central France and one of the most sparsely populated areas in the country. This level of specificity proved important. It allowed us to conduct targeted research, identify relevant local challenges, and refine our proposal around a real context rather than a broadly European one.
Research: Desk Work, Interviews, and AI
With a defined user group and a specific region to focus on, we moved into research. This involved both desk research and interviews, which together helped us build a clearer understanding of the problem.
Interviews allowed us to hear directly from people affected by the issue. This type of primary research is central to service design because it helps uncover needs, frustrations, and behaviors that secondary sources often miss (Stickdorn & Schneider, 2011). Even within the limited timeframe of a hackathon, conversations with users challenged some of our early assumptions and helped us ground the concept in real experiences.
Desk research provided a broader understanding of transport systems, demographic trends, accessibility challenges, and existing solutions. AI played a useful role during this phase. We used tools such as ChatGPT and Gemini to help synthesize information, identify relevant sources, organize findings, and strengthen aspects of the business proposal. In a time-constrained environment, having tools that could rapidly process and structure information enabled us to work more efficiently.
Liedtka (2018) argues that design thinking works because it combines user understanding with iterative problem-solving. The combination of interviews, desk research, and AI-supported information gathering helped us move through this process more quickly while keeping the solution grounded in real user needs.
Prototyping as a Tool for Clarity
With research completed and a clearer concept emerging, we moved on to creating personas. Personas are commonly used in service design to transform broad user groups into specific representations of the people being designed for (Stickdorn & Schneider, 2011). In our case, they helped us visualize the daily realities, transport challenges, and needs of elderly people living in rural Creuse.
Brown (2008) describes prototyping as a core component of design thinking because it enables teams to externalize assumptions and learn through experimentation. Buxton (2007) extends this idea by arguing that prototypes are not simply representations of solutions, they are tools for exploration that help teams discover what a solution should become.
This was exactly our experience. As we began building the prototype, guided by our personas and research findings, we were forced to answer practical questions that had remained unresolved during discussion. What exactly does the solution do? Which features are essential? How does the user interact with it? Building the prototype made these questions unavoidable and ultimately helped sharpen the concept.
One of the most important lessons was that prototyping is not simply a way to present an idea. It is a way to think through an idea. The prototype became a shared language that aligned the team and transformed abstract concepts into something concrete and understandable.
How AI Supported the Prototyping Process
In a hackathon setting, the ability to iterate quickly is essential. This is where AI tools became particularly valuable. We used ChatGPT and Gemini to support ideation, structure content, develop user flows, and refine elements of the prototype.
The greatest benefit was speed. Tasks that might normally take hours could often be completed in minutes, allowing the team to focus on improving the solution rather than getting stuck on execution. AI supported rapid iteration, making it possible to test multiple approaches within the limited timeframe of the event.
Liedtka (2018) highlights that design thinking succeeds because it creates opportunities for continuous learning and feedback. AI supported this process by lowering the barriers to experimentation. Rather than spending significant time developing a single version before seeking feedback, we could create, test, revise, and improve concepts much more rapidly.
However, AI did not generate the solution for us, nor did it replace the need for critical thinking. Instead, it reduced the effort required to visualize ideas, structure information, and create early versions of concepts. This allowed the team to spend more time discussing the problem, evaluating alternatives, and improving the solution. In that sense, AI supported innovation by accelerating the process rather than defining the outcome.
The Value of a Working Prototype for the Pitch
When it came time to present and pitch our idea, the working prototype became the centerpiece of the presentation. Rather than asking judges to imagine how the solution might address transport and accessibility challenges, we could demonstrate it directly.
The impact on communication was immediate. Judges and participants were able to engage with something tangible, ask specific questions, and provide focused feedback. Several improvements to the concept emerged directly from these conversations—insights that would likely not have surfaced if the idea had remained theoretical.
Another important learning was the value of failing early. Some of our initial assumptions did not survive the feedback process, and that was beneficial. Through conversations with coaches, participants, and judges, we identified weaknesses and areas for improvement that were not obvious when the idea existed only in discussion. Because we had a prototype, these issues became visible early enough for us to make meaningful changes.
Norman (2013) emphasizes that technology should support human-centered design rather than replace it. This became clear throughout the hackathon. AI made building faster, but it was human feedback that improved the solution. The prototype created the conditions for that feedback to occur.
Conclusion
Participating in the hackathon was an experience that extended far beyond the event itself. Working alongside people from different universities, countries, and professional backgrounds made the process richer and more challenging. Different perspectives encouraged new ways of thinking and pushed the solution in directions it might not otherwise have taken.
One of my biggest takeaways was the importance of narrowing the focus. At the beginning, it was tempting to think broadly and try to solve the problem for everyone. However, the more specific we became, from a European transport challenge, to elderly users, to accessibility, and finally to a specific region in France, the easier it became to research the problem, build personas, and create a meaningful prototype. The experience showed me that narrowing the scope does not reduce innovation; it helps create solutions that are more realistic and actionable.
AI proved to be a valuable tool in a time-constrained environment. It kept iteration fast and helped us move from an idea to something testable without getting stuck. However, the real value came from the team, the research process, the feedback received, and the willingness to continuously refine the concept.
Before the hackathon, I viewed prototyping primarily as a way to present an idea. After the experience, I see it as a way to develop, challenge, and improve ideas. In future projects, I am likely to prototype earlier and more frequently, not because the tools demand it, but because this experience demonstrated that clarity emerges when ideas become visible, testable, and open to feedback.
References
Briscoe, G., & Mulligan, C. (2014). Digital innovation: The hackathon phenomenon. Creativeworks London Working Paper.
Brown, T. (2008). Design thinking. Harvard Business Review.
Buxton, B. (2007). Sketching user experiences: Getting the design right and the right design. Morgan Kaufmann.
Liedtka, J. (2018). Why design thinking works. Harvard Business Review.
Norman, D. A. (2013). The design of everyday things: Revised and expanded edition. Basic Books.
Stickdorn, M., & Schneider, J. (2011). This is service design thinking: Basics, tools, cases. BIS Publishers.
Leave a comment