Struct A.I LTD
Struct
Since
Solution Overview
for
Every real estate project begins with one question: what can legally be built on this parcel? Answering it properly means resolving overlapping national, district and local statutory plans, reading unstructured Hebrew planning documents, deciding which plan supersedes which per sub-area, and layering geological, infrastructure and environmental constraints on top.
Today that answer costs $23,000тАУ89,000 per parcel in planners, appraisers and geotechnical consultants, and takes 6тАУ16 weeks. A developer screens roughly 30 parcels to acquire 2, so the same bill is paid 28 times to hear “no”. The cost of a slow “no” is the largest and least-discussed inefficiency in early-stage development.
Worse is a wrong “yes”. A regulatory or geotechnical constraint discovered after land acquisition, or after design has begun, means redesign, delay or a write-off. Architects design on parcels nobody has fully checked, because fully checking one is prohibitively slow and expensive.
Our Solution:
Struct is a SaaS platform for real estate feasibility and regulatory analysis, sold as an annual subscription.
In practice: a user selects a parcel, or uploads a blueprint in DXF, DWG or IFC. Struct pulls every statutory plan and GIS layer applying to that parcel, and models trained on planning documents extract the numeric values from each. A resolution engine then decides which plan governs which sub-area where they conflict, producing one unified building-rights table.
From there the platform inflates those rights into a maximum buildable envelope on the map, allocates buildings by permitted use with an initial apartment mix, and attaches economics from comparable transactions and appraisal standards. It also reports the gap between existing and approved state.
The output is a go/no-go verdict with the numbers behind it, in minutes rather than weeks. As design progresses, the same engine re-checks the drawing and flags conflicts before they become rework.
How it works:
Existing tools start from structured zoning data. They only work where a clean, machine-readable zoning layer already exists, and they answer massing and pro-forma questions once the rules are known. Struct inverts that. Our innovation is producing the structure rather than consuming it: a multi-agent pipeline that reads unstructured statutory planning documents in their native form and converts them into a machine-readable, spatially resolved constraint model.
Classification & Use Cases
Lifecycle stage(s) addressed:
Core technology types:
Key use cases in practice:
Maturity & Traction
Stage
Regions
Main:┬а
Team
employee(s)
Impact & SDGs
Contribution to any of the 17 Sustainable Development Goals:
Digitising one of the least digitised industries. – 11: Accessibility, services and demographic fit surfaced before approval. – 8: Productivity тАФ minutes replace weeks of expert time. 12: Conflicts caught pre-design avoid redesign waste. – 15: Sensitive land screened out, zoned land used fully.
Main impact areas
Team & Organization
Founding team:
Noam Talmor, CEO тАФ Computer Science and Business Administration; owns technical architecture, infrastructure and product. https://www.linkedin.com/in/noam-talmor-proptech/
Haim Ozeri, CTO тАФ Computer Science and Economics; leads the extraction and resolution pipeline. https://www.linkedin.com/in/haim-ozeri-9432a9349/
Michal Raznoshik, CPO тАФ Civil Engineering and Geology (dual degree), M.Sc. Geotechnical Engineering; owns the domain logic. https://www.linkedin.com/in/michal-raznoshik-377424263/
Advised by the former Deputy Director General of a national planning authority (Product Director), a PhD in urban planning and civil engineering, and a planning-policy specialist.
