Introducing AI into HVAC does not begin with sensors — it begins with what you already have: folders, drawings, invoices, Google Drive. We digitize them, structure them with AI and build a digital twin of your plant. On a real European reference aquatic center, three binders were digitized and the entire pool was captured — enough to operate the plant as a twin and plan a new one. Presented anonymously.
Finding the right contract clause, spotting a billing error, locating an as-built drawing — before a single new sensor, AI that can read and search your existing documents already pays for itself. It is also the honest first step: you see your plant clearly before you change it.
Binders, as-built drawings, invoices, Google Drive, e-mails — whatever documents the plant. Nothing is thrown away.
Paper and PDFs are scanned and read by AI (OCR) — meter readings, setpoints, equipment lists, contract terms become machine-readable.
A local knowledge base (RAG) makes everything searchable in plain language: „which contract clause covers the night tariff?“, „where is the as-built of the filter room?“
The structured data becomes a living model of the plant — one you can operate against today and plan a renewal, expansion or repowering on tomorrow.
Ask in plain language across folders, PDFs and Drive — the AI finds the clause, the drawing, the reading.
Meter points, setpoints, equipment and contract terms are pulled out and structured automatically.
Billing and contract inconsistencies surface — the documented savings before any hardware change.
The twin stays current as new readings and documents arrive — one model of the plant.
Renewal, expansion or repowering decisions are made on a real model, not on guesswork.
It runs on-premise — operating data stays in the building (GDPR by design, Art. 25/32).
Next: once the plant is a twin, we introduce AI where the savings are highest — the energy hogs. See Applications →
Documentation, a digital twin, then AI where the savings are highest — on your existing plant. Single property, portfolio or chain.
Request intro call →