Functional DemoAI Platform

Home Intelligence OS

Homeowner-first AI platform for exterior home condition, storm exposure, and consent-based contractor matching.

What it solves

Homeowners lack a trusted, non-salesy way to understand their home's exterior systems, storm exposure history, and maintenance priorities. Existing tools are lead-generation-first and treat the homeowner as a sales target rather than the primary user. There is no independent place to plan lifecycle, assess risk, and estimate financial impact without pressure.

Built for Homeowners (primary); contractors, inspectors, insurance-adjacent partners, and property managers as consent-gated secondary users.

Overview

Home Intelligence OS helps homeowners understand, plan, protect, and improve the exterior systems of their home over time. It analyzes exterior condition, interprets storm exposure history, evaluates lifecycle status, scores risk, and estimates financial impact, presenting AI analysis with explicit confidence levels and separating it from verified inspection findings. A consent model lets homeowners connect with qualified contractors only when they choose to, supported by a lead marketplace, a contractor matching engine, and admin tooling. The platform is built around homeowner trust rather than a sales funnel.

What this demonstrates

A multi-package monorepo with separate homeowner, contractor, and admin portals plus an API, backed by a model-provider abstraction that defaults to Anthropic and degrades to a deterministic stub. It implements domain engines for risk scoring, storm exposure, financial impact, explainability, and contractor matching over a PostGIS-enabled Postgres model, with JWT auth, rate limiting, and a consent/disclosure boundary enforced in code.

What makes it different

The platform must feel like a trusted advisor, not a sales funnel — a homeowner-first intelligence layer for the exterior life of the home, where consent is a hard system boundary.

Homeowner-firstEvery product, data, AI, and business decision prioritizes homeowner trust, control, clarity, and benefit.
Trust is the most important product featureA homeowner can use it without feeling manipulated, pressured, sold to, or exploited.
Consent is a hard boundaryExplicit, granular, revocable, and auditable — enforced at the API, database, workflow, and contractor-access layers.
Separate AI from verified findingsAI analysis is clearly distinguished from verified professional inspection findings.
Confidence-aware and non-alarmistNo fear-based language, no exaggerated urgency, no implied damage without evidence.
No hidden lead generationNo default opt-ins for contractor sharing; the marketplace never becomes the hidden center of the product.
Explainable findingsEvery conclusion carries reasoning, supporting inputs, confidence, and known limitations.

Core engines · 18

Detection Normalization Engine
Ingests and normalizes raw exterior detections into structured findings.
Lifecycle Assessment Engine
Estimates lifecycle stage for roof, siding, gutter, window, door, and accessory systems with confidence.
Risk Scoring Engine
Scores risk across weather, age, material, condition, and maintenance factors — explainable and non-alarmist.
Insurance Qualification Engine
Provides educational insurance guidance without guaranteeing coverage or eligibility.
Financial Impact Engine
Models repair and replacement ranges, maintenance budgeting, and deferred-maintenance risk as estimates.
Storm Exposure Engine
Interprets historical hail and wind exposure mapped to the property — never presented as confirmed damage.
Home Report Orchestrator
Assembles the versioned, downloadable homeowner report from all engine outputs.
Report Section Generator
Generates the individual report sections from engine findings.
Explainability Engine
Produces reasoning, supporting inputs, confidence, and limitations for each conclusion.
Maintenance Roadmap Engine
Builds short, medium, and long-term maintenance plans with priorities and cost ranges.
Recommendation Decision Engine
Generates conservative, confidence-aware recommendations such as monitor, document, or schedule inspection.
Contractor Matching Engine
Matches contractors by trade, service area, quality, and specialty — only after explicit consent.
Homeowner Profile Engine
Manages homeowner accounts, properties, and preferences.
Homeowner Engagement Engine
Drives homeowner engagement flows and notifications.
Homeowner Intelligence Dashboard Engine
Powers the exterior intelligence dashboard over time.
Lead Marketplace Engine
Consent-based opportunity creation, eligibility filtering, and lead lifecycle with data minimization.
Platform Administration Engine
Admin operations: user, property, contractor, report, and consent management.
Workflow Orchestration Engine
Coordinates asynchronous, retry-safe, idempotent workflows across engines.

How the risk score works

Risk score

Weather exposure
Storm and climate exposure affecting the property.
Age indicators
System age relative to expected lifecycle.
Material vulnerability
Susceptibility of the exterior materials involved.
Visible condition signals
Condition indicators drawn from detections.
Maintenance gaps
Deferred or missing maintenance.
Regional climate factors
Local climate patterns raising or lowering risk.
Historical event patterns
Past storm and event history for the area.
Detection confidence
Confidence in the underlying detections.

How it works

Homeowner

Add a property by address
Ingest and normalize exterior detections
Interpret storm exposure history
Assess lifecycle and risk, confidence-aware
Model financial impact
Generate an explainable report and maintenance roadmap
Grant explicit consent to share
Receive consent-based contractor matches

Contractor

Complete business verification (license, insurance)
Define service area and trade categories
Receive a homeowner-consented opportunity
Accept or reject the matched opportunity
Communicate through approved workflows
Submit outcome updates and maintain quality score

Administrator

Monitor AI outputs and flagged findings
Verify contractors
Audit consent records
Resolve disputes
Enforce policies

Key screens

Property dashboard
Exterior intelligence dashboard
Report viewer
Storm history timeline
Lifecycle overview
Risk explanation page
Financial planning page
Maintenance roadmap
Contractor consent flow
Contractor match results
Privacy settings
Admin portal

Data model

Homeowner
Contractor
Property
Exterior Detection
Normalized Finding
Storm Event
Lifecycle Assessment
Risk Assessment
Financial Assessment
Report
Recommendation
Maintenance Plan
Consent Record
Marketplace Opportunity
Contractor Match
Workflow Run
AI Output
Audit Event

How it is built

A pnpm/Turbo monorepo: Next.js/React frontends for each audience, a TypeScript API service, shared domain, schema, AI, and observability packages, a Drizzle + Postgres/PostGIS database, and Redis for caching, rate limits, and queues, all runnable via Docker Compose.

APIs and services

Anthropic

Classification

Industries: Real EstateHome Services
Capabilities: AI Automation Authentication Analytics Reporting Document Generation Geospatial Analysis Role-Separated Interfaces Multi-Tenant Architecture
Technologies: Docker Next.js Node.js PostgreSQL React Redis Tailwind CSS TypeScript Drizzle Zod

Building something with similar AI Automation, Authentication, Analytics functionality?

Start a project and reference this one — the conversation begins with what you have already seen.

Start a project