Independent principal — technical & data-architecture direction for data platforms
I design and direct data platforms institutions rely on — and the AI applications that put them to work. Twenty-five years of public-data systems: the data architecture behind OpenCoesione.gov.it (Italy's national portal for EU cohesion funds), Openpolis' integrated data platform, the CIPE deliberations archive. Since 2024: LLM-based monitoring in production, natural-language access to data at prototype stage — an AI application over data is only as good as the platform beneath it, and I design both layers.
Directing now: the MAPS data-lake for Fondazione Gran Sasso Tech · technical director, Fondazione Openpolis · CTO, DEPP.
What's public here. Most of what I direct runs in private, client-owned infrastructure. The pinned repositories are the public traces — the public-data lineage, the delivery method, and where the mathematics started.
geojson-italy— GeoJSON/TopoJSON boundaries for every Italian municipality, region and province. Public-data plumbing others build on.django-popolo— the Django reference implementation of the Popolo open-government data standard (people, organizations, posts).tmetric-minimal-mcp— a Model Context Protocol server; a small public piece of the AI-over-data work.guglielmo.github.io— source of guglielmo.celata.com.formule-in-movimento— animated videos making maths and physics visual. I came to AI from the mathematics; this is that thread.txmd— a small terminal Markdown viewer, from everyday work.
Elsewhere: LinkedIn · guglielmo.celata.com