Real estate & proptech

Real-estate data is duplicated across portals with inconsistent attributes, agencies run on spreadsheets and messengers, tenant requests, contracts and invoices are handled by hand, and a valuation still depends on one expert’s opinion. The regulatory load is rising: estate agents are obliged entities under the EU anti-money-laundering framework (Directive (EU) 2015/849), and the recast Energy Performance of Buildings Directive (Directive (EU) 2024/1275) will push energy data into every listing and transaction. Eurostat’s house price index (2024) shows how quickly markets move, which is exactly why manual valuation and stale listings cost money.

• Listings duplicated across portals with inconsistent data
• Agencies run on spreadsheets and messengers
• Tenant requests, contracts and invoices handled by hand
• Valuations depend on one expert’s opinion

We build listing portals and marketplaces with search, maps and lead capture, agency CRMs with deal pipelines, property and tenant management systems with payments and maintenance requests, and mobile apps for agents and tenants. Search runs on a proper index with geo queries, listings are normalised into one schema on ingestion, and documents such as contracts and invoices flow through a queue-based pipeline with human approval steps. Personal data of buyers, tenants and applicants is handled to GDPR (2016), including retention limits and access rights.

• Listing portals and marketplaces with search, maps and lead capture
• Agency CRM and deal pipelines
• Property and tenant management systems, payments and maintenance requests
• Mobile apps for agents and tenants

AI moves real-estate numbers in valuation, data quality, documents and lead handling. McKinsey (2023) estimated that generative AI could generate $110–180 billion or more in value for the real-estate industry, mostly in customer interaction, document handling and analytics. Automated valuation models give a defensible second opinion next to the expert; listing deduplication and enrichment turn ten inconsistent copies into one record; Document AI extracts terms from contracts and invoices; lead scoring and enquiry assistants answer the first questions and route the rest. Valuation models used for lending decisions fall under EU AI Act (2024) transparency expectations, so we ship them with explanations.

• Automated valuation models
• Listing deduplication and data enrichment
• Document AI for contracts and invoices
• Lead scoring and assistants for enquiries

Because deduplication, enrichment and Document AI are patterns we have already shipped in retail, and real-estate data has the same problems with different nouns. Sixteen years and 500+ projects have taught us to normalise data on the way in, to keep a human approval step wherever money or contracts move, and to treat every portal feed and payment API as unreliable by default. Security verification for portals that hold applicant documents is covered under security services; see also cases.