The platform

One autonomous AI pipeline, from raw building data to verified result.

Struxiva's fully automated AI reads all your building's scattered documents and data and turns them into one structured, validated dataset. Get the data right, and everything else just works.

Struxiva isn't a dashboard you have to feed. It's an AI engine that ingests the data you already have, validates it, and runs detection, prediction and verification continuously, so the answers are waiting for you, not waiting on you.

Why AI is the foundation, not a feature

The intelligence is the product.

Most building software shows you data and leaves the thinking to you. Struxiva inverts that. The AI does the unifying, the fault-finding, the forecasting and the savings math, automatically and continuously. People step in to decide and act, not to wrangle spreadsheets.

The five stages

1 · Ingest

Struxiva pulls in BIM/IFC models, HVAC diagrams, technical drawings, energy audits, renovation reports, IoT and time-series data, and primary-/secondary-side energy consumption, from whatever tools and formats you have today.

2 · Validate

The AI cleans, structures and cross-checks every source into one continuously updated model of the building. Messy, fragmented and legacy data is the expected starting point, not a blocker.

3 · Detect

It surfaces operational defects, energy inefficiencies and hidden underperformance in real time, and localizes them, telling you which system, which floor, which unit.

4 · Predict

It forecasts what will fail and when (remaining useful life, failure forecasting) and how the building should run (energy use and peak demand), so you act before problems and costs arrive.

5 · Verify

Every action is measured against a clean baseline, turning "we think it helped" into auditable, realized savings.

Talk to your building

On top of the pipeline sits a conversational AI interface. Ask a question in plain language, "Why is the east wing overheating?", "What's driving Building C's bill?", and get an instant, data-backed answer drawn from the full validated model. Anyone on the team can use it.

Works with what you already run

Struxiva is a software layer, not a rip-and-replace. It connects to your existing systems and data sources, and improves the more data it sees.

Architecture

Every document goes through the same durable pipeline: stored and deduplicated, filtered before any AI is called, classified, routed to the extraction method that matches its format, reconciled against every other document describing the same building, and materialized into one graph. That graph, not any single document, is what the rest of the product reasons over.

Struxiva ingestion and knowledge-graph architecture Documents are uploaded, stored and deduplicated, filtered by a zero-cost triage step, classified and extracted using a method matched to the document format, reconciled through entity resolution, and materialized into a Postgres graph. The graph feeds the detect, predict and verify engines and can be exported as BRICK RDF. Upload Multipart or signed URL Storage & dedup Cloud storage, content-hash check Tier-0 triage Zero-cost filter, no AI spend yet Classify & extract Method matched to document format Entity resolution Reconciled across every document Graph (Postgres) Brick-classed, durable & queryable Detect · Predict · Verify Product engines BRICK / RDF Export, on demand

Security & hosting

  • Runs on Google Cloud infrastructure — Cloud Run for processing, Cloud Storage for documents, Postgres for the structured graph.
  • Every document is scoped to its organization and building at the storage-path level, so tenants' data is isolated by design.
  • Uploads travel over short-lived, signed URLs directly to storage, encrypted in transit.
  • Processing runs as durable, checkpointed workflows: if a worker fails mid-document, work resumes from the last completed step rather than being lost.

Full details on data residency, retention and our data-processing terms are available on request — ask us and we'll walk you through them.

Integrations

Struxiva ingests the formats building teams already have:

IFC PDF XLSX / XLS CSV DWG / DXF Images

The validated graph is also exposed via API, so it can feed existing BMS, CAFM or ERP systems instead of living in a silo.

Have a specific BMS, CAFM or ERP system in mind? Tell us what you run and we'll confirm compatibility.

Book a demo