DIS2026X1 Hackathon
DAY 3; MAY 8 | 10:00
Build a Robot AI Agent in 2 Hours & Win HP ZGX Nano AI Station!
At DIS2026X1, we are hosting a focused, high-intensity hackathon designed for developers who want to move beyond theory and build something real. Each team will have two machines connected to each other for even more performance and ability to work with more or larger AI models, and a strict two-hour window to design and deploy an AI agent that controls a physical robot system.
Date | May 8
Location | Kistamässan
Duration | 4 hours
*exclusively available only to Data Innovation Summit delegates
A Real-World Deployment Hackathon
Most hackathons prioritize ideas over execution or spend valuable time on setup and infrastructure.
This format is intentionally different.
You will work directly on pre-configured systems, using local AI hardware designed for fast iteration and real-world deployment. The focus is not on pitching concepts, but on building a functioning agent under real constraints.
There is no extended onboarding, no complex environment setup, and no room for overengineering. The challenge is to think clearly, build efficiently, and deliver within a fixed timeframe.
The Challenge
Participants will develop an AI agent that interacts with and controls a physical robot. The system supports multiple input modalities, including visual and audio signals, allowing teams to explore a range of interaction patterns and behaviors.
The objective is not only to make the system work, but to push it as far as possible within the constraints of the hardware and time available. Teams are expected to make deliberate trade-offs, optimize performance, and design agents that behave in a coherent and responsive way.
Full challenge details will be provided at the start of the session.
Format and Schedule
The hackathon will take place on May 8th from 10:30 to 14:30 and follows a tightly structured format.
The session begins with a 30-minute introduction, where the challenge and environment are presented. This is followed by a two-hour build phase, during which teams develop their solutions. After submission, there is a short break while entries are evaluated, concluding with a 30-minute award announcement.
This compressed format is designed to prioritize focus, execution, and clarity of thought.
Evaluation
Projects will be evaluated by a small jury consisting of two human judges and one AI-assisted evaluation component.
The assessment is based on three core dimensions:
– Performance, measured by how effectively the team utilizes the available hardware
– Use of sensors and inputs, including how well different modalities are integrated
– Agent quality, including responsiveness, coherence, and overall behavior
The goal is not only technical correctness, but the ability to create something that feels complete and intentional within the given constraints.
What You Will Work With
Each team will be given access to two ZGX Nano AI Stations, configured as one combined machine giving twice the performance and VRAM in advance to allow immediate development. The environment is designed to remove friction and enable rapid prototyping, so teams can focus entirely on building and testing their solutions.
A physical robot system will be available as part of the challenge, enabling direct interaction rather than simulation-only development. The setup supports both local execution and hybrid approaches, giving teams flexibility in how they structure their solutions.
Who Should Apply
This hackathon is intended for conference attendees with an intermediate level of experience in development or AI.
Participants should be comfortable working under time constraints and making practical engineering decisions quickly. Prior experience with robotics is not required, but a willingness to experiment and adapt is important.
Teams may consist of a minimum one and up to three people.
We encourage participants to apply with a team in advance.
Application
Participation is limited to 15 teams. Applications will be reviewed and accepted on a rolling basis until all spots are filled.
If you are attending DIS2026X1 and want to test your ability to build under pressure using real AI hardware, this is a direct way to do it.
Apply now to secure your place.