I design and build data platforms, consolidate heterogeneous sources and apply traceable AI for public administrations and companies. I speak the language of technical teams and of the people who make the decisions.
Technical and product capabilities for R&D&I projects, pilots and public procurement of innovation, focused on open, interoperable and auditable solutions.
I don’t stop at the technical side: I translate the needs of business and public-service teams into data products people actually use. I work with users and decision-makers to prioritise, agree on indicators and goals, manage the backlog and explain results in language everyone understands.
Lakehouse architectures and ELT pipelines on GCP, AWS and Azure. Data quality, governance and cataloguing from day one.
Unifying data from multiple digital analytics providers into a corporate platform: one model, consistent metrics and reliable dashboards.
AI-generated reports and documents, RAG over your own documentation and agents — always with human validation, traceability and EU AI Act compliance.
Sensor networks and monitoring (energy, indoor climate, space usage) connected to data platforms that produce actionable recommendations.
Fast proofs of concept and pilots for open innovation challenges and public procurement of innovation, from idea to working prototype.
Professional displays for hotels and venues: integration, data-driven dynamic content and LED installations.
A selection of recent work.
Integration of data from several digital analytics providers into a regional government's cross-departmental data platform, with a common model and analytical exploitation.
ELT platforms, product and marketing analytics (Piwik PRO, Segment) and dashboards for high-traffic real estate portals.
A network of meters and LoRa links on Raspberry Pi to monitor a temporary power grid at a large event, with real-time dashboards.
Integration of professional hotel displays and an 80-metre networked LED installation with dynamic content.
Scoped projects, verifiable deliverables and no black boxes.
Understand the challenge, existing data sources and the indicators that matter.
A working proof of concept in weeks, with real data.
Controlled deployment, impact measurement and tuning with users.
Industrialisation, documentation, knowledge transfer and guaranteed reversibility.