Hackathon · AI and data
GLOBAL: exploring housing policy with data and sources.
We developed GLOBAL during the 2026 URV–T-Systems AI Hackathon. It is an observatory prototype that connects housing legislation, market indicators and documentation to investigate public policies through a common interface.
The approach supports comparison and source-backed questions without presenting exploratory scenarios as proven causal effects.
- Context
- URV–T-Systems AI Hackathon · 2026
- Field
- Housing policy
- Status
- Hackathon prototype
The problem
Comparing housing policies means interpreting laws and data from different regions and periods. Economic and social variables interact, and no single chart can explain the whole picture.
What we built
The prototype connects a conversational interface with documents, indicators and comparative views. It separates information retrieval, data processing and explanation generation so sources can be reviewed.
Its architecture combines Next.js, FastAPI, analytical databases and retrieval-augmented generation (RAG).
My contribution and scope
I worked with the team to build the product and turn heterogeneous sources into an experience we could demonstrate during the event.
GLOBAL is a technology exploration and comparison prototype. It is not a validated tool for causal estimation or autonomous public decision-making.
Sources and links
Public references and project material documenting its scope.