AI agents hackathon winners

🏆 25+ teams spent their Friday at Qonto's Paris office, hacking on agent-focused projects in all shapes and sizes. Special focus on the winners. If you’re curious about the finalists as well, here’s a quick Linkedin post. And you can find at the end of this post the full video of the 6 finalists.

A shorter version of this -with news and links about and around Dust- was first published on our newsletter. Best way to stay tuned is to follow us right there!

Most impactful use case (1st place): Simon Cleriot, Khaled Khebbeb, Jarek Ecke and David Aparicio.

First and foremost, who are you?

Jarek: I'm Jarek, working as a GenAI Engineer here in Paris and originally from Germany. I love the dynamic and impact AI has in this city and believe it's the best place right now to be in Europe to build.

Khaled: If I could choose a word to qualify myself, it would be "passionate". I'm fond of philosophy, cinema, and music. I've been playing guitar and bass for over 15 years now, I've been part of some bands with a main focus of having fun and giving tiny shows in bars. I'm also a big fan of video games, and those brought me to IT.
I hold a PhD in computer science with a focus on self-adaptive systems, I've been working on the Cloud-related fields for a few years now.
Now, I spend all of my time as principal cloud architect optimizing processes and strengthening our infrastructure at Comeen.

David: I'm David, Software Engineer graduated from INSA in Lyon and UNICAMP (Brazil). I have worked for Facebook on the MIT AppInventor project as a graduation project. I did a 2-year thesis where I discovered Docker, Kubernetes and DevOps. Then I worked for big players like AMADEUS (2 years) and OVHcloud (5 years). I'm now working as Senior DevSecOps at Sopht, the Green ITOps tool to reduce your IT carbon footprint.

How would you describe your experience during this day? Was it your first hackathon?

Jarek: Great organisation, good food and very qualified teams. Also really nice support by the Dust team. We have a quite active sharing of events in my company - everything GenAI related gets sent to me. It was my first Hackathon in a while, loved to participate when I was still a student.

Khaled: Absolutely positive. Our hosts (you!) were so classy and welcoming. The community was open-minded and easy-going. We enjoyed every tiny bit of interaction we had. The organization of the event itself was a non-fault. Thank's and cheers for that.
We use Dust at Comeen, so I've joined the Slack community and subscribed to Dust's news. Then, as soon as the announcement was made, we subscribed to participate.

David: Before the quick but in-depth Alban's workshop the same morning, I had no idea what Dust had to offer. He did a wonderful job. It was a truly outstanding event, meeting exceptional people with extraordinary backgrounds and contagious energy.
I didn't count my hackathon participation, but I think I'm slowly approaching ten. My last prize was in 2014, at Facebook São Paulo, my team won an honorable mention for the project that used Facebook as a tool to find missing people.

Can you describe what you built?

Jarek: We built an agentic workflow that lets non-tech users fix code bugs automatically end-to-end by just describing what is not working (the rest is handled by the workflow).

Khaled: I'd call it the "headache saver"; an enabler for non-technical staff who have to maintain or interact with some tech.
We took the example of brand managers, PMs, marketing officers who might need to maintain some kind of "brochure" or company website with static contents. But it could be applied on any use case where the definition of "what should be" is simple enough to be detailed in a knowledge base or "KB" (like a Notion documentation) and where the definition "what is going wrong" is tangible. One might call these descriptions as "functional specs" or "functional rules".

So starting from that, we worked on a multi-layered autonomous bug fixing approach, allowing anyone to interact with a chatbot that has knowledge of the KB and that can process the bug report. The next step would be sending a summary to an agent that has the ability to pull, analyze and write on the codebase to directly propose a fix to the reported bug.
On the demo, we showcased a direct fix, with automatic release of the changes to production. But since we can write operations, the right path would be: opening a pull request to propose the fix, and deploying that change to some temporary and disposable environment for testing and QA.

David: Imagine a world where non-technical team members can seamlessly contribute to development projects without waiting for tickets to move into the sprint. Our Dust Agent bridges the gap between the two teams. Let's break down the silos!

Why did you decide to work on this?

Jarek: It's a real problem that is quite easy to solve in a PoC [proof of concept] but is not solved at scale yet.

Khaled: The impact here is time saving and it also enables autonomy. Non-tech staff would no longer require a busy dev to be available, and the dev won't need to delve into the KB to understand the specifics. They would instead just review a PR and save on both headaches and keystrokes. In this scenario, the non-tech person gains enthusiasm and a sweet feeling of accomplishment. Every person that has ever experienced this kind of interaction -i.e., the simple bug that takes months weeks to be fixed- knows the frustration.

David: This comes from an observation we've made in all the companies we work with. Non-tech people can't develop, they don't know our tools (like Cursor, Windsurf or Roo Code) so they have to create a ticket, wait, wait and wait for that ticket to leave the backlog to get into a sprint and try a quick resolution.

How did this day change your perspective on AI?

Jarek: It was a great experience seeing how easy it became on a small scale to build agentic workflows. This trend will go up to bigger use cases!

Khaled: I've mostly been a consumer of LLMs, mainly for optimizing my organization, saving time on some tasks, exposing my reasoning on some problems, looking for cognitive biases on my approach. I've also been building some personal projects that consume some AI APIs to implement some fun ideas, but never did I really delve hands-on into those MCP/agentic considerations.
Dust's Hackaton definitely opened my eyes on this world of possibilities. And especially, I realized this is something really fun, satisfying and affordable enough to get started on building real stuff.

David: The ease of the tool! A few weeks later, I had a company seminar with a three-hour workshop on AI, which I had to lead. In just a few dozen minutes, I had three professional use cases to demonstrate the value of AI: a technical documentation expert from Notion, an assistant who lists customer problems from Slack, and finally, an agent who summarizes what's happened since we've been away from the office (very handy for retrieving important information after a company seminar in the French Alps). This enabled the startup's employees to come up with ideas for the last remaining hour. Thank you Dust for this valuable hand!

Most impactful use case (2nd place): Axel Darmouni.

First and foremost, who are you?

I am Axel Darmouni, currently an AI engineer consultant at Sia AI for more than 2 years working on GenAI topics. I am passionate about the topic: publishing on LinkedIn for market and model related updates, and using more my Twitter account as a personal research blog related to the topic.

How would you describe your experience during this day? Was it your first hackathon?

I heard about the hackathon from the Dust LinkedIn post! I have been long interested in what Dust was proposing, and this hackathon was an opportunity to get to know more the people of Dust while honing my Data & fullstack skills under pressure.

It was my second one: my first was a hackathon organized by Entrepreneur First with Meta and Hugging Face, related to the use of small models, which was a blast! I ended in the top 7, out of 22 teams, and mine was the only of the group which had tried to tackle a research work over the week-end.

Can you describe what you built?

I built a product to help sick people get quicker to their nearest pharmacy.

To tell you the truth, when I am sick, I am usually feeling too bad to even budge a muscle: I’d rather stay comfy in my bed than face the pain and get out of bed. The origin of this idea was to ask myself the question: “how would I make myself move?”. I am usually a thought-to-action kind of person, so my answer was that “I need a goal. Even if it’s not a complete goal, I need a clear goal and for the process to be as fast as possible”. Therefore, I was going for it at first to create this app for myself. Given that it was a hackathon and I had low experience in App Development, I settled for a Web App in React instead. The two pushes I needed were the following:

• Medication idea: Taken from the French “ANSM” database of over-the-counter medication
• Identification of nearby pharmacies: Use of Google Maps API

The idea that I had in mind was also to test several tools I had under my radar but couldn’t find time to test:

• Using Structured Outputs with an Enum list for classification, which was donesuccessfully for diagnosis identification extracting a taxonomy from the ANSM database
• Getting a feel of the Gemini models: I used to be a Claude enjoyer, but this vibe test allowed me to validate their strong performances
• Using an MCP, more precisely the Maps MCP, which in the end ended being slightly less successful than making a workflow related to the maps API

Using Cursor for Frontend Development, which definitely converted me as a cursor fan: I even took the cursor subscription after this hackathon!

Going alone into this hackathon was both intentional and unintentional at the same time: I first registered with two friends of mine, who told me three days before that they could not make it. I asked myself whether or not I wanted to take a challenge and go for a project idea alone, and so I did! Which ended up being quite the pressure, but quite fun as well.

How did this day change your perspective on AI?

The day was a really good time to me: a very high quality organization, and a presentation in front of the jury was more than I could have asked! What I am most excited about, after this hackathon, is obviously the given Dust subscription as a reward. I’ve been itching to try the tool for personal use, and I think it will quite satisfy me! Overall, this hackathon was a great experience: can’t wait for the additional ones to come!

Most creative use case: Gabriel Landman and Gary Klajer.

First and foremost, who are you?

Gabriel: Hey Im Gabriel Landman, student engineer in CentraleSupelec. I love Bilinear Algebra, Franz Kafka, Muay Thai and my 11th arrondissement. Currently working at Scoreplay in Lisbon as a software engineer.

Gary: I’m Gary Klajer, a mathematics student with a builder mindset. My academic journey spans EPFL, École Polytechnique, and Université Paris-Saclay. With a strong inclination toward applied research, I’ve worked on privacy-preserving methods for medical LLMs at Dauphine PSL and am currently interning at Owkin, focusing on explainable AI (XAI) for histopathology. Next year, I’ll join the MRes MVA (Health Track) program at ENS Paris-Saclay and plan to pursue a PhD thereafter.

How would you describe your experience during this day? Was it your first hackathon?

It was pure madness--in the best way possible! It was a very short timeline, but it forced us to go at full speed. It was also amazing to meet and learn from inspiring founders and mentors.

We came across the Dust AI Agents Hackathon through a friend who always has the latest infos about tech events (no one knows how he does it).

But it wasn’t our first hackathon.

Can you describe what you built?

Our project, DebateGab, is a game-like platform designed to enhance debating skills. It enables users to engage in debates and having AI agents providing feedback and insights to refine their argumentation skills.

Why did you decide to work on this?

We found the idea of agents representing not “roles” but actual persons amusing, so we started thinking about a creative use case leveraging conversations with real or fictional characters. We are aware it won’t revolutionize the tech industry but was super fun to work on !

People found it fun, and several people asked us to publish it, so we extended the project and plan to make it accessible to a larger audience. But you can already try it out (on laptops only) right now (here). Use the credentials (newsletter, 123456), or create an account and enter an OpenAI API key if you want to have your own account.

We can't wait to hear your feedback!

How did this day change your perspective on AI?

We relied super heavily on coding assistants like Gemini 2.5 Pro, and we would have never been able to do it so quickly without it. It made us realize the power of such AIs when used by someone who is already comfortable with the technology and is capable of organizing and orchestrating the project properly.

Thanks to both Qonto and Qdrant for helping us shaping the day. Thanks to all members of the jury – and of course, every one who came with a big smile, and limitless energy.

🍿 You can watch below the demo of the 6 finalists. Gary and Gabriel start at 1min45 — Axel starts at 28min24 — Simon, Khaled, Jarek and David start at 47min33. Enjoy 🍿