Restb.ai is a business-focused computer vision and AI platform operated by Mercurio Platform, S.L. Its services analyze real estate images and related data for room and feature tagging, condition and quality signals, visual similarity, captions, photo-compliance flags, generated property descriptions, watermark and duplicate detection, comparable-property workflows, and appraisal-complexity assessment. These outputs are machine predictions derived from submitted images and data. They are not a physical inspection, appraisal, survey, title record, code review, insurance determination, fair-housing approval, or guarantee that a listing statement is complete and accurate.
What Restb.ai does
Customers integrate Restb.ai through cloud APIs and contracted products. The property-description service says it combines listing details, image-derived signals, location data, tone, and style controls to generate remarks. The broader platform serves MLSs, portals, search and marketing businesses, appraisers, inspectors, investors, iBuyers, and insurers. Product access, call limits, inputs, outputs, service levels, geography, languages, and commercial rights depend on the service order and current contract.
Restb.ai's terms say clients retain ownership of uploaded visual content but license it for service delivery. Clients are responsible for content rights, accuracy, account security, lawful use, end-user notices, and sufficient protection. The terms state that outputs and services are provided with significant accuracy and suitability disclaimers, and that customer data may include photos and metadata. They identify Restb.ai as a processor for that data and name AWS infrastructure in EU and US locations. Contract terms should be reconciled with current security and data-processing documents.
Who it is for
Restb.ai may fit organizations processing property imagery at scale, including MLSs, portals, valuation teams, appraisal firms, insurers, investors, and real estate software vendors. It is not a consumer home inspection, licensed appraisal, underwriting decision, property-condition warranty, accessibility certification, copyright clearance, or replacement for agents and qualified reviewers responsible for final records and communications.
What to verify before adopting
Define each model, version, label, confidence threshold, supported property type, geography, image requirement, update process, and human-review rule. Test representative images against recent inspections, appraisals, source listing data, prohibited-content policies, and known edge cases involving poor lighting, staging, renovation, damage, duplicates, people, personal belongings, and inaccessible areas. Review false positives, false negatives, bias, correction, drift, monitoring, audit logs, latency, rate limits, outages, and fallback. Confirm image and metadata rights, sensitive-data restrictions, retention, deletion, subprocessors, transfer locations, encryption, access, incident response, DPA, SLA, output rights, and downstream disclosure. PropAIdir has not independently tested Restb.ai or reproduced vendor claims about model accuracy, consistency, compliance, languages, cost reduction, valuation improvement, security, scalability, or business outcomes.
