House Made Ready Limited
Version
1.0
Date
September 2026
Prepared By
Luke Roberts (Founder & CEO)
Classification
Confidential
Governance Standard
UK DSIT AI Risk Management Toolkit & HM Treasury Orange Book Framework
House Made Ready Limited ("House Made Ready") is a UK-based property intelligence platform that integrates real-time planning data, including Permitted Development Rights (PDR), Article 4 directions, and Conservation Area constraints, with HM Land Registry price-paid records and broader property datasets. The platform automatically generates 60-second property scorecards and data cards that power its proprietary AI engine, Maggie AI, to produce automated, accurate architectural insight packs, financial ROI projections, Strategic Staging Reports, Investment & Mortgage insight packs, development feasibility studies, and real-time conversation for B2C and B2B users.
This report establishes House Made Ready's formal AI Risk Management Framework in alignment with the UK Department for Science, Innovation and Technology (DSIT) AI Risk Management Toolkit. It defines the governance structures, systematic risk assessments, mitigation controls, and continuous audit protocols necessary to manage operational, legal, technical, and ethical risks across all active and pipeline AI capabilities.
Maggie AI Core Intelligence Engine & Data Card Framework
Proprietary AI engine that ingests real-time HM Land Registry price-paid data, Local Planning Authority (LPA) portals (PDR, Article 4 directions, Conservation Areas), and geospatial mapping to build structured property data cards. These cards drive the automated generation of:
Estate Agent Marketing Suite
Automated content generation engine utilising OpenAI GPT-6 to craft property descriptions, social media collateral, portal listing optimisation, and buyer communication templates.
Professional & Trade Matching Logic
Algorithmic matching system that pairs high-intent homeowners and developers with verified architects, surveyors, and builders based on geographic data, property constraints, and project scope.
Generative Staging & Visual Design Engine
Generative image modelling (integrating tools such as DALL-E / Midjourney) to produce automated visual staging overlays and preliminary design concepts for Strategic Staging Reports.
Inbound Conversational AI (Voice & Text)
Automated conversational agents for scaling incoming customer inquiries, onboarding estate agents, SME property developers and homeowner leads, and assisting initial property triage.
House Made Ready adopts a multidisciplinary governance approach to ensure comprehensive oversight across technical performance, legal and regulatory compliance, data privacy, and commercial operations.
Executive Lead & AI Officer
Luke Roberts (Founder & CEO)
Ultimate accountability for overall AI risk appetite, sign-off on high-risk model deployments (including core updates to Maggie AI), and enforcement of emergency system suspension protocols.
Commercial & Partner Lead
Jacqueline Butler (Commercial Growth Lead)
Oversight of B2C homeowner privacy data & B2B estate agent, developer, and trade network data compliance, collection and routing of user feedback loops, and validation of AI-generated marketing and sales output accuracy.
Technical & Data Pipeline Lead
Lead Architect / Technical Partner
Oversight of data ingestion security across HM Land Registry and Local Planning Authority (LPA) APIs, real-time data card generation integrity, API rate-limiting, Maggie AI algorithm drift monitoring, and prompt engineering guardrails.
Legal & Compliance Advisor
External Counsel / DPO
Safeguarding UK GDPR compliance, conducting copyright/IP risk reviews for AI-generated outputs, reviewing third-party AI vendor SLAs/DPAs, and maintaining alignment with ICO and DSIT guidelines.
Risks are categorised according to the four core pillars of the UK GOV.UK AI Risk Management Toolkit: Legal & Regulatory Compliance; Technical Robustness & Data Quality; Fairness, Bias & Societal Impact; and Accountability, Redress & Remediation.
Third-Party LLM Data Leakage: Users or estate agents inputting PII or confidential seller details into GPT-4-powered tools, exposing data to model retraining without Enterprise Data Protection Agreements (DPAs).
Maggie AI Planning & Feasibility Hallucination: Maggie AI misinterpreting LPA datasets (PDR, Article 4 directions, Conservation Area boundaries), resulting in inaccurate Property Data Cards and flawed Architectural Insight Packs.
Financial & Mortgage Data Drift: Stale HM Land Registry price-paid data or flawed API calls causing Maggie AI to output inaccurate ROI projections or misleading Investment & Mortgage Insight Packs.
Generative Visual & Content Infringement: Generative design tools (DALL-E/Midjourney) or automated copy engines producing copyrighted imagery or deceptive representations in Strategic Staging Reports.
Algorithmic Trade & Professional Referral Bias: The matching engine systematically favours specific architects, surveyors, or builders due to skewed review weighting or postcode demographic factors.
Unvetted Voice/Text AI Customer Binding: Conversational agents hallucinating pricing, quoting incorrect fee structures, or making unauthorised contractual commitments to B2B or B2C users.
R-01: Data Privacy & Third-Party LLM Security
R-02: Maggie AI Planning Guidance Hallucinations & PDR Errors
R-03 & R-05: Financial Valuation Drift & Data Pipeline Accuracy
R-04: Generative Staging & Visual/Content IP Safeguards
R-05: Algorithmic Trade & Professional Referral Bias
R-06: Automated Voice & Customer Interaction Safeguards
House Made Ready enforces a strict vendor vetting procedure before integrating external APIs, LLMs, voice agents, or spatial data feeds. This framework ensures all third-party integrations comply with UK DSIT standards and UK GDPR requirements.
Data Privacy & Retention
Vendor must contractually guarantee zero data retention for training purposes. Must support UK/EU data residency where applicable.
Verification: Review API Terms of Service, Enterprise DPA, and Data Transfer Assessments.
System Reliability & SLA
Vendor must guarantee operational uptime and sub-second API latency response times to maintain the 60-second scorecard delivery commitment.
Verification: Status page monitoring & SLA agreement.
Cybersecurity Standards
Mandatory SOC 2 Type II or ISO/IEC 27001 certification. End-to-end encryption required for all property records and PII (TLS 1.3 in transit, AES-256 at rest).
Verification: Security compliance documentation review.
Model Explainability
AI model providers must furnish comprehensive system documentation detailing dataset provenance, known hallucination rates, safety alignment filters, and prompt injection defences.
Verification: Technical model cards & system documentation.
Operational & Financial Resilience
API providers must maintain clear deprecation policies (minimum 6 months advance notice) and demonstrate operational stability.
Verification: Review developer documentation, vendor roadmap disclosures, and financial health assessments.
House Made Ready operates a three-tiered continuous monitoring and auditing framework to ensure Maggie AI, property data card generation pipelines, and third-party API dependencies maintain high levels of accuracy, safety, and legal compliance.
Tier 1: Real-Time Automated Logging & Telemetry
Tier 2: Monthly Human QA & Deliverable Sampling
Tier 3: Quarterly Governance & Regulatory Review
In the event of a critical failure, House Made Ready initiates a 3-step escalation protocol:
Step 1: Containment & System Fallback (Immediate)
Step 2: Diagnostic Investigation (Within 2 Hours)
Step 3: Remediation, Statutory Reporting & Model Update
Report Sign-Off & Approval
Approved By: Luke Roberts, Founder & CEO, House Made Ready Limited
Next Review Date: December 2026
End of Report
Transforming properties through data-driven insights and architectural expertise.
House Made Ready Limited
Co. No: 16933702Company number: 16933702