No One Else Has This
Innovations where a thorough patent search and competitive analysis found no equivalent anywhere in the market or patent landscape.
Three-Agent Adversarial Architecture
Buyer's Rowan represents exclusively the buyer. Seller's Rowan represents exclusively the seller. A Transaction Rowan sits between them with strict information walls — each agent sees only what its party should see. No commercial platform deploys this configuration. The nearest patent (Microsoft, 2016) covers only two-party negotiation and was abandoned. AdHome's mediator + information wall concept is unpatented territory.
Materials Term Sheet → Auto Purchase Agreement
A plain-language deal summary that both parties negotiate in human-readable terms — which then automatically transitions into a legally binding purchase agreement as each clause is agreed. Combines ADR (Alternative Dispute Resolution) frameworks with AI document generation. No platform offers this neutral, progressive transformation from handshake to contract. Opendoor's contract patent covers templating, not this live-transition mechanism.
Unified Scan → Appraisal → Virtual Tour → Digital Twin Pipeline
A single seller-guided smartphone scan simultaneously produces four outputs: an AI property appraisal, virtual showing assets, disclosure documentation, and a persistent digital twin that lives with the home through ownership. Individual pieces exist (Restb.ai, Matterport, Zillow AVMs) but nobody chains all four outputs from one guided scan. The digital twin persisting through ownership is the most novel element — Matterport's 84 patents cover 3D capture, not this unified pipeline.
Transaction-to-Ownership Lifecycle Continuity
Every existing platform stops at closing. HomeZada starts after purchase. Homebot is B2B-only. AdHome is the only system that spans listing → transaction → closing → ongoing ownership as one continuous platform. The digital twin is the retention mechanism that keeps homeowners on-platform between transactions (which happen only once every 7-10 years). No patent covers this cross-phase lifecycle continuity.
Transaction Narrative "Consciousness"
A system that maintains holistic, narrative-level awareness of the entire transaction — knowing not just the state of each document and task, but the emotional arc, the history of communications, the reasoning behind every decision, and the likely next issues. Unlike a state machine (Opendoor's patent), this is contextual intelligence that understands the deal as a story. USPTO patent search returned zero prior art. File immediately.
AI Disclosure Auto-Generation from Scan Data
California AB 723 (Jan 2026) created new AI disclosure requirements. AdHome can automatically generate seller disclosure documents by analyzing property condition data gathered during the guided scan — cross-referencing against California's extensive disclosure checklists and flagging every required disclosure item. TrueHouse AI and others analyze existing disclosures; no one generates them from scan data. No blocking patent found. Strong regulatory tailwind.
Pieces Exist. The Combination Is New.
These innovations draw on proven patterns from other industries — but no one has applied the combination to real estate transactions.
AI Escalation-and-Learning Loop in Real Estate
Customer service AI that escalates to humans and learns from every escalation is a well-known pattern (Intercom, Zendesk, etc.). Applying it to full real estate transactions — where the AI handles what it can, escalates gracefully to licensed human experts when it can't, and uses every escalation as training data — is entirely new to this industry. Only 27% of Americans trust AI for high-stakes financial decisions. This model bridges that gap rather than pretending it doesn't exist.
Trojan Horse GTM Strategy
REX Real Estate attempted a frontal assault on agents — spent $115M and permanently closed. Compass invested $1.6B making agents stickier. AdHome's middle path: onboard agents as power users of the platform, learn from their behavior to train AI that eventually handles their tasks, then transition consumers directly. The business gets revenue and training data during Phase 1, and a clear path to disintermediation in Phase 2. More sophisticated than anything tried before in proptech.
CarMax-Style Lending Inside Transaction Flow
CarMax's insight was that embedding lending inside the car buying experience — rather than as a separate step — increased conversion, loyalty, and margin simultaneously. LendingTree and Bankrate do standalone mortgage comparison. Rocket's integration with Redfin is agent-mediated. AdHome embeds lending comparison and pre-approval directly into the transaction flow at the moment of peak intent — when an offer is being drafted. This contextual timing is the differentiator, and it's proven in adjacent industries.
Two-Sided Active AI Matching
Most matching in real estate is passive — buyers search, listings sit. AdHome deploys two active AI agents: the buyer AI continuously refines and seeks matching properties; the seller AI continuously refines and seeks matching buyers. They're both hunting simultaneously, informed by behavioral signals that neither party explicitly stated. The only relevant patent (Copley, 2014) was abandoned in 2015. Zillow's search patents cover passive search functionality, not bilateral proactive seeking. This bilateral-active model is unpatented in real estate.
Potential Future Innovations
High-potential concepts to layer in as the platform matures. Ranked by impact and feasibility.
♻ Smart Contract Escrow
Propy proved it works: escrow completed in 6 minutes, closing in 3 days, 60% cost reduction. Propy secured a $100M facility in early 2026 to scale this. AdHome could integrate smart contract escrow as the settlement layer for every transaction. Regulatory complexity varies by state but California has been receptive.
🎲 Predictive Negotiation Intelligence
Each party sees probability distributions — Monte Carlo simulations showing the likelihood their offer is accepted, what counteroffer is likely, how long the negotiation will take, and what the final price distribution looks like given comparable transactions. Converts guesswork into informed strategy. Addresses the information asymmetry that agents currently exploit.
🔌 IoT-Connected Digital Twin
Connect the digital twin to IoT sensors (HVAC, water, electrical) for real-time property health monitoring. Predictive maintenance reduces homeowner costs 10-30%. Creates a continuous data stream that improves property valuation accuracy and generates recurring subscription revenue. Phase 3+ — hardware integration complexity is high.
🌐 Community Intelligence Network
Anonymized, aggregated data from homeowner digital twins creates hyperlocal market intelligence unavailable anywhere. Which neighborhoods have the highest deferred maintenance? Where are HOA disputes clustering? What's the real condition of homes in a ZIP code vs. what listing photos show? Institutional buyers, insurers, and lenders would pay premium for this ground-truth data layer.
🛡 Embedded Insurance (Dynamic Pricing)
The digital twin has ground-truth property condition data that no insurer has today. Dynamic insurance pricing based on actual property state — not zip code averages — benefits good-condition homeowners and creates a revenue stream ($500-$2K per homeowner/year recurring). The embedded insurance market is projected to reach $950B by 2030 (35% CAGR). Partnership model with carriers; AdHome provides the data layer.
💎 Property Tokenization for Equity Access
Allow homeowners to tokenize a portion of their equity for liquidity without selling or refinancing. The digital twin's verified condition data provides a credible collateral basis for tokenization. The market hit $10B+ in 2025 but faces regulatory uncertainty, particularly around SEC classification of property tokens. A Phase 3+ play contingent on regulatory clarity.
Patent Strategy
Based on full USPTO landscape research across all nine AdHome innovations. The 2025 USPTO guidance shift now favors AI patent eligibility — timing is favorable.
Tier 1 — File Immediately (Provisionals)
Tier 2 — File Within 3 Months
Tier 3 — File Within 6 Months
🚫 Do Not Patent
Patent vs. Trade Secret Recommendations
| Innovation | Recommendation | Rationale |
|---|---|---|
| Three-Agent Architecture | Patent | Competitors will observe this from product behavior; cannot keep secret once live |
| Term Sheet Auto-Transition | Patent | Visible in user experience; fully reverse-engineerable from the interface |
| Property Scan Pipeline | Hybrid | Patent the unified pipeline; trade-secret the specific ML models and training data |
| Disclosure Auto-Generation | Patent | Process is visible to users; regulatory environment demands transparency |
| Negotiation Simulation Engine | Hybrid | Patent the architecture; trade-secret the simulation algorithms and weighting |
| Two-Sided Active Matching | Patent | Observable from product behavior the moment the platform launches |
| Transaction Narrative System | Hybrid | Patent the architecture; trade-secret the prompting and context management techniques |
| Progressive Qualification Model | Trade Secret | Internal UX logic; difficult to patent under 101; not easily reverse-engineered |
| Trojan Horse GTM | Trade Secret | Business strategy; unpatentable; competitive advantage from execution, not concept |
Key Competitor Patents to Watch
Zillow Group
US10,375,306 and US10,530,997 (building interior data capture from mobile devices) are the highest-risk patents for AdHome's scan feature. The core Zestimate patent was significantly weakened by the PTAB. Multi-faceted search patents may affect matching.
High Risk — Scan FeatureMatterport (CoStar)
94.4% grant rate. Broad coverage on 3D data capture, scene alignment, room classification, and property layout generation. CoStar acquired Matterport in 2025. Mitigation: emphasize AI-driven condition assessment over spatial reconstruction. Note: Matterport joined LOT Network.
High Risk — Digital TwinOpendoor
Siamese network AVM patents (US11,869,045; US11,481,818) cover AI valuation. State model patents (US12,190,051) cover transaction state machines. Pending contract generation application (US20200034940A1) could affect document workflow. All distinguishable with careful design.
Moderate Risk — ValuationRocket Companies
Core patent US11,921,696 covers event-listener architecture for mortgage document processing. Focus is entirely on mortgage processing and underwriting — not the buying/selling transaction. Not a current concern unless AdHome enters mortgage origination.
Low Risk — Different DomainISG / Bot Mediation
ISG's US 12,572,606 (granted March 2026) covers AI-to-AI contract negotiation — general domain, not real estate. Bot Mediation has a provisional patent on AI mediation for legal disputes (not real estate). Both are distinguishable but require monitoring as they evolve their claims.
Low Risk — DistinguishableJoin LOT Network Immediately
LOT Network (5,800+ members) provides automatic cross-licensing protection against patent assertion entities (patent trolls) — who file 60%+ of all US patent lawsuits. Matterport is already a member. Membership is free for companies under $25M revenue and covers access to ~6 million patent assets (24% of all US patents). This is the single highest-ROI defensive measure available before any product is built.
What Should We Call the AI?
34 candidates researched, 18 eliminated for conflicts. Top 5 recommendations below.
Italian "trovare" (to find). Coined word = fully ownable. Premium sound.
Musical term + acronym (Automated Real estate Intelligence & Assistance). Elegant.
Shakespeare = sanctuary. Gender-neutral. Clean TM space.
Trust + home. Signals honesty. Strategic differentiator vs agents.
AdHome Name Availability
Domain, trademark, and app store availability for the AdHome brand as of March 2026.
Domain Availability
Trademark, Company & App Store Checks
US Proptech — Clear
No existing companies using the "AdHome" name in US property technology or real estate. The name is available in this space.
USPTO — "AD-HOME" Trademark
An "AD-HOME" trademark exists in telecommunications (International Class 38). This is a different class from real estate technology services. Risk is low — trademark classes are separated, and the mark does not cover proptech or software services.
Apple App Store — Clear
No "AdHome" app currently exists in the Apple App Store. The name is available for iOS app registration.
Google Play Store — Clear
No "AdHome" app currently exists on Google Play. The name is available for Android app registration.
Name availability reflects research conducted March 2026. Domain prices and trademark status should be re-verified before any purchase or filing. Consult a trademark attorney before filing USPTO applications given the existing AD-HOME mark in Class 38.