From 6 to 8 November, Turku (Finland) becomes the meeting point for more than a thousand AI builders, who will spend 72 hours at Since AI 2026 working on carefully scoped challenges brought forward by companies, research groups and public bodies. Participation is free, there is a €50,000 prize pool, and the judging rule is that only working demos count. SinceAI is backed by Bayer, Sandvik, Valmet and a range of European VCs. The organizers have invited FIWARE to shape its own challenge track within that program, and we want to build it with you.

The track is built around problems where interoperability genuinely matters, approached through Context Brokers, NGSI-LD, and Smart Data Models, and using open, synthetic, or sandbox data so that teams can build something real without coming anywhere near sensitive production systems. The themes on the table include smart-city operations and mobility data, energy communities, water and environmental monitoring, industrial monitoring and predictive maintenance, and digital twins.

What makes this track different from a standard AI challenge is the underlying stack. FIWARE is not just an interoperability layer, it is a complete, proven pipeline from data collection to AI deployment: IoT Agents and Context Broker collect and manage real-time context data; FIWARE Draco routes and transforms it; Cosmos GE stores the historical datasets that feed model training. The result is structured, semantically rich NGSI-LD data that can be used directly to fine-tune open-source AI models — from Hugging Face or the EU’s own AI-on-Demand platform — for a specific sector. Teams at Since AI 2026 will not just be building interoperable systems; they will be building sovereign AI: applications where the data pipeline, the model, and the deployment environment are all open, auditable, and under the control of the organization running them, not a third-party platform. This is the architecture the EU AI Act and Gaia-X data spaces are designed to enable, and FIWARE is the open-source stack that makes it practical today.

The FIWARE community has already demonstrated this in practice: TinyML models deployed on edge devices for urban mobility cyberphysical systems, AI-powered digital twins built on linked open data, Smart Data Models used as the schema layer for flood-risk and Earth observation AI, and European sovereign AI infrastructure. SinceAI 2026 is the opportunity to take that body of work into the hands of a new generation of builders working on real challenges.

The track is yours to shape. We are inviting FIWARE community members and iHub organisations to bring forward challenge briefs. If you have a real problem where interoperable, sovereign data matters — in your city, your industry, your organisation — this is the format to test it: 72 hours, international teams, working demos, and an eight-week follow-up path for the strongest results. Write to us at opportunities@fiware.org by 8 October if you want to put a challenge forward or discuss what that would look like.

What comes out of it:

• Several international teams working on a problem of yours in parallel, each leaving with a working prototype
• Reusable outputs — NGSI-LD integrations, Smart Data Model extensions, trained models, reference architectures, open repositories
• A new generation of AI builders who understand that sovereign, interoperable data is the foundation serious AI applications are built on

How you can be involved:

• As a challenge owner — bring a real problem from your organisation or city
• As a mentor, architecture reviewer or judge — on site in Turku or remotely
• As a community connector — join our FIWARE Discord and link up directly with Since AI’s 10,000+ builder network; we are bridging both communities so FIWARE experts are reachable by builders before, during, and after the event
• Alongside the strongest teams through the eight-week follow-up period, aimed at pilots and partnerships

Developers who would rather take part on their own can register directly at sinceai.ai/hackathon, free of charge.

“FIWARE provides the foundation for open and interoperable smart systems. Since AI can bring a concentrated community of builders to apply that foundation to difficult real-world problems.”