The Slovak Football Association gained a new way to work with more than a decade of data and project know-how. We built an MCP solution that connects internal documentation with live data, allowing authorised users to ask questions in natural language, run analyses and build new digital services on top of existing information.
The Slovak Football Association began digitising its data and processes in 2010. By the time this project started, its digital ecosystem covered approximately 360,000 members, 1,700 clubs, 44 football associations and 5,000 teams, while processing results from around 2,200 matches every week.
Over the years, this growing volume of data was accompanied by extensive project documentation describing how the systems work, internal processes, data structures, relationships between individual data points and the rules governing their use.
SFZ relies on this information when preparing conferences, creating materials for regional associations and clubs, communicating with partners and supporting internal decision-making. Producing the right output, however, required finding information across several parts of the system, combining it and validating it against the project documentation.
The goal was therefore to make both data and know-how accessible from one place. Users needed to be able to ask a question in natural language and create analyses, comparisons, presentations, manuals or materials for partners and football associations without preparing database queries or manually searching through documentation.
For SFZ, we built an MCP server on top of its project documentation, processed using RAG. This gives AI the context it needs to understand how the systems work, how internal processes are structured, where data is stored and how individual data points relate to one another.
When working with current data, the solution connects to a MongoDB MCP server. The context provided by the documentation helps the model determine where the required information is located, how different data points are connected and how they should be used correctly.
Users can simply ask a question in natural language through their AI tool. The model combines relevant information from the documentation with available data and can produce an analysis, comparison, presentation material, documentation-based answer or even a draft user manual.
The AI can work only with the data, documents and functions made available through the individual MCP tools. SFZ therefore retains control over which sources the model can access, while outputs can be verified against the original data or documentation.
Because MCP is an open standard, the same prepared sources can also be used by different compatible AI tools. Every new task can build on the existing connection between data, documentation and project context rather than starting from scratch.
Following implementation, the SFZ team began using the solution for analyses and comparisons, preparing materials for conferences, football associations and partners, searching project documentation, creating user manuals and validating new ideas against real-world data.
One of its first practical applications was the preparation of statistical materials for a conference organised by the Nitra District Football Association. The first infographic was created in under an hour, including validation, while adapting it for another football association took approximately ten minutes.
Using the same data and AI foundation, the SFZ team then independently built the public Slovak Football Statistics portal. Its first version was created over a single weekend.
At launch in 2026, the portal covered:
The portal makes historical statistics on matches, goals, cards, attendance, clubs and registered players, coaches and referees easily accessible. SFZ, regional associations and clubs can use the data when preparing presentations, communicating with partners, comparing developments over time and making decisions about future activities.
The fact that the SFZ team was able to build the portal independently demonstrates that connecting data, documentation and AI creates value far beyond faster access to information. It gives the client a foundation on which they can build their own outputs and launch new digital services as new needs emerge.
SFZ now has a shared foundation for both everyday work with data and future AI use cases — from internal analysis to solutions designed for clubs, football associations and the wider football community.
“The bart.sk team feels like an extension of our own team, and their strong understanding of both sport and sports management made it easy to identify where AI could bring real added value. This solution has given us much greater freedom when analysing and working quickly with complex data across the Sportnet.online platform, well beyond match results alone. Bringing raw data, project context and AI together gives us a powerful tool for tasks that, until now, had been practically out of reach due to the time and cost involved.”
