AI-assisted evidence for the built environment

Where a site or a project stands, with every finding sourced.

Luma Intelligence is an AI-assisted evidence engine for the built environment. It shows the people who fund, approve and design buildings where a project stands on the qualities that Europe's housing and renovation money now depends on: reuse, bio-based and circular materials, climate adaptation, social cohesion, inclusion and accessibility.

See the two demos Talk to us

Prototype. Every output is decision support: a screening validated by the planner, or a pre-assessment with human sign-off. It is never a verdict or a certificate. The architect stays the author.

One engine, two products

From where to build, to how good the scheme is

1 · Site intelligence

Where can we build? in use in Slovenia

"Where in Vransko can we build an apartment building with ten apartments?" The answer is a list of ranked candidate sites, each with its source and confidence, then a site card: transport, schools and services, slope and orientation, what the site permits. Flood and heat layers follow by M3.

2 · Project intelligence

How good is the scheme? prototype

Upload a brief, sketches or drawings. Seven evaluators return cited findings, ranked moves and blind spots. A quick assessment through one or two lenses is the entry tier.

Its exports, a pre-filled NEB self-assessment, a quality statement for tenders and competitions and a funding-readiness gap list, are drawn from the same review. The exports are planned for 2027.

Architects use it and stay the authors. Housing funds, municipalities, developers and lenders pay for the evidence. Verified solution providers and green financiers take part under a published neutrality charter: recommendations are never for sale.

Two demos · prototype

The two ends of one market, on the same engine

Demo A · Site finder

Site intelligence

"Where in Vransko can we build an apartment building with ten apartments?"

Ranked candidate sites on a map. Each line has its source and confidence, and each site has a short assessment whose citations are checked, with unverifiable ones flagged. Screening, validated by the planner.

Open the demo

Demo B · Project intelligence

A quick assessment at the entry price

A renovation project, two lenses: bio-based and circular materials, renovation and reuse.

Quick assessment, two lenses: what a €10 ticket buys when larger clients co-fund the engine. Ranked moves come with impact, effort and confidence. The checker checks every citation and flags the ones it cannot verify. The full seven-lens review is the larger product. The demo uses a fictional project.

Open the sample report

One minute

Selma Špendl, founder and CEO

1-minute pitch video (coming)

Proof so far:

A paying client in Slovenia for the first location study
A working prototype, built in-house, that checks every citation and flags the ones it cannot verify
A team that builds AI agent systems itself and works in spatial planning and engineering

Housing fund, municipality, developer, lender or practice?

Talk to us about the site finder, a project review or a pilot in 2027.

pilot@lumaintelligence.eu