Start/End
User Action
Decision
System
AI Agent
Outcome
Data
Legal/Compliance
Vendor
Money
Complete Buyer Journey
Every step from first visit to moving in — with progressive qualification, AI matching, and full transaction management.
Phase 1: Discovery & Qualification
Buyer visits AdHome
Browse listings freelyNo signup required — zero friction
AI tracks browsing behaviorTime spent, photos viewed, features clicked, neighborhoods explored
Create account (email only)Save favorites, get notifications
Buyer's Rowan assignedBegins learning preferences — explicit + implicit
AI engages transparently"I noticed you keep looking at single-story homes near good schools..."
Set explicit criteriaPrice range, bedrooms, location, must-haves
Want to engage with sellers?
Progressive qualification kicks in
Phone verification + basic financials
Pre-qualificationAdHome internal assessment + lending marketplace pre-approval
Buyer is qualifiedCan now request showings and make offers
Phase 2: Search & Matching
Buyer's Rowan actively searchesScans all listings against explicit criteria + learned preferences
Receives outreach from seller AIsActs as gatekeeper — filters spam, surfaces genuine matches
Match scored & rankedLocation 25% | Price 20% | Features 20% | Condition 15% | Lifestyle 10% | Timing 5%
Review matchesAI explains why each property was recommended
Interested in a property?
Request showing
Phase 3: Showings
Showing type?
Virtual Tour3D model from seller's scan data — available instantly
Self-guided explorationAI highlights features, flags concerns
Lockbox AccessVerified buyer gets access code
Self-guided in-personAI provides pre-showing prep package
AdHome ContractorIndependent showing agent assigned
Guided in-person tourContractor answers questions, takes notes
Post-showing feedbackStructured feedback + passive behavioral signals + contractor observations
AI refines matchingFeedback loops into preference model for better future matches
Phase 4: Offer & Negotiation
Ready to make an offer?
Buyer's Rowan develops strategyAnalyzes comps, market velocity, seller's listing history, competition
Buyer sets parametersMax price, preferred terms, contingencies, flexibility
Buyer's Rowan submits offer to mediatorOnly shares what buyer approves — information wall in effect
Transaction Rowan facilitates negotiationAsync by default — AI negotiates 24/7. Sync option available.
Counter-offer?
AI advises on each counter"Seller countered at $1.32M. Market data suggests $1.28-1.30M is fair. Here are your options..."
Terms agreed
Materials term sheet generatedNeutral, plain language, both parties understand without an agent
Phase 5: Contract Through Closing
Purchase agreement auto-generatedIn-house counsel review. AI explains every clause in your language.
Review & sign on your own timeNo artificial urgency
Automatic transition
Escrow opensEarnest money collected via secure wire
AI orchestrates in parallel:Inspections | Lending | Title search | HOA docs | Contingencies
Inspections scheduledGeneral, pest, roof, sewer, foundation — from vetted vendor network
Issues found?
Repair negotiationSame negotiation platform, same agents, full context
Lending finalizedAI recommended lender + marketplace alternatives with pros/cons
Title cleared, docs preparedWire fraud prevention active — multi-factor verified
Final walk-through
Closing dayFunds transferred securely
Keys in hand. Home ownership begins.Digital twin active. Maintenance, warranties, insurance managed.
Complete Seller Journey
From deciding to sell through receiving funds — with AI-powered valuation, marketing, and negotiation.
Phase 1: Property Scan & Valuation
Seller decides to list
Seller's Rowan assignedGuides entire process, speaks seller's language
Guided property scanPhone photos to LiDAR — whatever you can do
Gaps in coverage?
No gapsProceed to AI analysis
Vendor fills gapsProfessional photography, 3D scan, measurements
One scan creates three outputs
AI processes scan dataComputer vision analyzes every photo for features, condition, quality
Property ValuationRecommended price + range + reasoning + improvement suggestions with ROI
Virtual Tour Assets3D model, photo gallery, interactive walkthrough
Digital Twin CreatedLiving record: rooms, fixtures, appliances, condition, history
CA Disclosures auto-generatedTDS, NHD, lead paint — from scan data. Seller reviews & attests.
Phase 2: Listing & Marketing
AI generates marketing packageListing description, photo selection, staging advice, timing recommendation
Seller reviews & approvesInteractive collaboration — refine together. Nothing goes live without approval.
Set desired termsPrice, contingencies, closing timeline, showing preferences
Listing goes liveAdHome internal + Zillow, Redfin, Realtor.com, MLS (concurrent)
Seller's Rowan actively seeks buyersSearches qualified buyer AIs, contacts them through agent-to-agent protocol
Performance dashboardViews, inquiries, showing requests, market position, AI recommendations
Phase 3: Showings & Offers
Showings coordinatedSeller sets preferences: times, notice period, showing types allowed
Offers received?
Multiple offers?
Single offerStandard negotiation flow
Multi-offer biddingAI ranks by price, terms, contingencies, buyer strength
Disclose other offers?Seller decides — AI advises on strategy & legal bounds
Seller's Rowan negotiatesMaximizes value within seller's parameters
Terms agreed → purchase agreement
Escrow → inspections → closingAI manages everything. Seller stays informed.
Sold! Funds received.
Three-Agent Negotiation Architecture
How the buyer's Rowan, seller's Rowan, and neutral mediator interact — with information walls and full transparency.
BUYER
Shares goals privately"Max $1.3M, need 3 bed, 30-day close ideal"
Buyer's RowanDevelops strategy. Analyzes comps, market, seller listing history.
AI advises: "Offer $1.25M"Buyer approves, adjusts, or asks for alternatives
AI submits to mediatorOnly approved terms shared — wall enforced
Mediator echoes backBuyer sees exactly what mediator is working with
Counter from seller?AI advises: "Seller wants $1.32M. Data suggests $1.28-1.30M fair."
Buyer decides next move
Sees agreed termsPlain language. Fully understood.
INFORMATION WALL
MEDIATORNeutral
INFORMATION WALL
SELLER
Shares goals privately"Want $1.35M, flexible on close date, no repairs"
Seller's RowanDevelops strategy. Knows market position, days on market, competition.
AI advises: "List at $1.35M"Seller approves, adjusts, or asks for alternatives
AI responds to mediatorOnly approved terms shared — wall enforced
Mediator echoes backSeller sees exactly what mediator is working with
Counter to buyer?AI advises: "Buyer offered $1.25M. Counter at $1.32M to anchor."
Seller decides next move
Sees agreed termsPlain language. Fully understood.
Materials Term SheetNeutral document — both parties understand without an agent
Purchase Agreement Auto-GeneratedIn-house counsel review. Sign on your own time.
Key principles: Each Rowan can be as aggressive as it wants within legal bounds. The ONLY constraint is the information wall. The mediator only works with what's explicitly shared and echoes everything back so both sides can verify. All negotiation types (initial deal, repairs, contingencies) use the same platform with full context.
Negotiation Modes
Asynchronous (Default)
AI agents negotiate 24/7 on behalf of parties. Key decision points surfaced when human input is needed. Each side responds on their own time. No scheduling nightmare.
Synchronous (Optional)
Both parties can go live for real-time negotiation. AI facilitates the exchange. System asks if parties want to switch to sync. Best for time-sensitive decisions.
Appraisal Evolution Strategy
How AdHome's property valuation capability evolves from a helpful tool to an industry-disrupting standard — phase by phase.
Legal context: The term "appraisal" has legal significance. Federally related transactions require USPAP-compliant appraisals by licensed appraisers. AdHome's AI output must be labeled "property valuation" or "estimated value" — never "appraisal." However, GSE appraisal waivers now cover 23% of loans and are expanding.
Phase 1 — Launch
Property Valuation Tool
No licensing required
- AI-generated "estimated value" from scan data
- Recommended price + range + reasoning
- Improvement suggestions with ROI
- Comps analysis, market data
- Used for listing guidance only
- Traditional appraisal arranged if lender requires
Phase 2 — Year 1-2
Hybrid Model
Partner with licensed appraisers
- AI does 90% of the work
- Licensed appraiser reviews & signs
- Like Automax.ai model
- USPAP-compliant output
- Accepted by lenders
- Each human review trains the AI
Phase 3 — Year 2-3
Licensed Subsidiary
Own appraisal licenses
- AdHome appraisal subsidiary
- Licensed appraisers on staff
- AI primary, human QA
- Faster turnaround, lower cost
- Building regulatory relationships
- Data moat deepens
Phase 4 — Year 3+
Regulatory Acceptance
AI-primary valuations accepted
- Work with GSEs on AVM standards
- Leverage expanding waiver programs
- Industry-leading accuracy from data moat
- Human oversight for edge cases only
- Set the standard others follow
How the Scan Gets Smarter Over Time
Seller guided scanPhotos, measurements, condition notes
AI extracts features & conditionComputer vision: fixtures, surfaces, appliances, quality
AI generates valuation estimate
Traditional appraisal also done?
Yes — compare resultsDelta between AI and human appraisal = training signal
AI learns from discrepancyWhat did the human see that AI missed?
No — track sale priceActual sale price vs. AI estimate = ground truth
AI calibrates accuracyOver/under estimation patterns detected
Model retrained with new dataEvery transaction makes the valuation more accurate
Property condition database growsGround-truth data insurers, lenders, investors pay for
Competitive Positioning
| Capability | AdHome | Zillow Zestimate | Automax.ai | HouseCanary | Traditional |
|---|---|---|---|---|---|
| Interior condition data | ✓ (scan) | ✗ | ✓ | ✗ | ✓ |
| Seller-guided (no pro needed) | ✓ | N/A | ✗ (inspector) | N/A | ✗ |
| Creates virtual tour | ✓ | ✗ | ✗ | ✗ | ✗ |
| Creates digital twin | ✓ | ✗ | ✗ | ✗ | ✗ |
| Auto-generates disclosures | ✓ | ✗ | ✗ | ✗ | ✗ |
| Lender-accepted | Phase 2+ | ✗ | ✓ | ✓ | ✓ |
| Improvement ROI suggestions | ✓ | ✗ | ✗ | ✗ | Sometimes |
| Cost to consumer | Included | Free | ~$300-500 | B2B only | $500-1,000+ |
The real advantage: Zillow's Zestimate has 7.1% off-market error because it has NO interior data. AdHome's guided scan provides ground-truth condition data that no other AVM has. This is the data moat — every scan makes the model more accurate, and the property condition database becomes the most valuable asset in the platform.
Property Scan Deep Dive
The guided scan process — from basic phone photos to comprehensive LiDAR. Scales to what the seller can do.
Scan Tiers
Basic
Phone Photos
Minimum to list. Room-by-room photos, exterior shots. Anyone can do this.
- Basic property profile
- AI valuation (wider range)
- Simple virtual tour
- Digital twin (basic)
Standard
Detailed Photos
Multiple angles, close-ups of fixtures, appliance photos, condition notes.
- Rich property profile
- AI valuation (tighter range)
- Interactive virtual tour
- Digital twin (detailed)
- Auto-generated disclosures
Comprehensive
LiDAR + Professional
Measurements, model/serial numbers, surface condition, 3D spatial data.
- Complete property profile
- AI valuation (highest confidence)
- 3D Matterport-quality tour
- Digital twin (comprehensive)
- All disclosures auto-generated
- Insurance-grade condition data
Key principle: Sellers can start basic and add detail over time. AdHome vendors fill any gaps the seller can't or won't cover. The system accepts whatever the user can contribute — never creates a barrier.
What the Scan Creates
Guided Scan Complete
AI Processing PipelineQuality check → Classification → Feature extraction → Condition scoring → Cross-photo correlation
Property ValuationPrice, range, reasoning, risks, improvements
Virtual TourPhotos, 3D model, interactive walkthrough
Digital TwinRooms, fixtures, appliances, condition, history
CA DisclosuresTDS, NHD, lead paint — auto-generated
Lending Flow (Annelys Thesis)
The loan manager coordination tool — acting as a virtual loan officer and general contractor. Same trojan horse pattern: tool first, compete later, replace eventually.
Buyer's Lending Journey
Buyer needs financing
AdHome internal pre-qualificationBasic financial assessment — learns accuracy over time
Lending marketplace activatedCarMax model: primary partner + competing lenders
AI gathers borrower docsW2s, tax returns, bank statements — organized and packaged
Multiple lenders competeRate, terms, fees — all presented side by side
AI recommends one lenderExplains why. Shows all alternatives with pros/cons. Buyer decides.
Buyer selects lender
AI coordinates underwritingGeneral contractor role: manages communication between all parties
The General Contractor Role
AppraiserSchedule, coordinate, track
LenderDocs, conditions, timeline
Title CompanySearch, insurance, recording
UnderwritingConditions, verifications
InsuranceHomeowner's policy
AdHome AI coordinates ALL of theseLike a general contractor managing subcontractors
Trojan Horse Evolution
Phase 1
Tool for Loan Officers
- Coordination dashboard
- Document management
- Timeline tracking
- System learns from every interaction
Phase 2
Compete
- Own pre-qualification
- Automated doc gathering
- AI handles routine tasks
- Loan officers handle exceptions
Phase 3
Own Underwriting
- AI runs underwriting analysis
- Human review for edge cases
- Licensed under own NMLS
- Competing with traditional LOs
Phase 4
The System IS the LO
- End-to-end automated lending
- Human escalation only
- Lowest cost in market
- Data advantage unassailable
Escrow Through Closing
Everything that happens after the purchase agreement is signed — managed automatically.
Purchase agreement signed
Automatic — no manual handoff
Escrow openedLicensed escrow partner activated
Earnest money collectedSecure wire with multi-factor verification — fraud prevention active
AI orchestrates everything in parallel
InspectionsGeneral, pest, roof, sewer, foundation
AI analyzes reportsCategorizes: safety / major / moderate / minor
Issues found?
Repair negotiationSame platform, same agents
LendingLender selected, docs submitted
AI manages conditionsTracks every condition to close
Appraisal needed?
Appraisal coordinatedThrough vendor network
Title searchAI flags liens, encumbrances
HOA docs reviewedCC&Rs, financials, assessments
Title insuranceProcured through vendor network
CA disclosures finalized
Contingency trackingDeadlines monitored. AI warns before deadlines hit.
Final walk-throughScheduled and coordinated. Verify condition and repairs.
Closing docs preparedIn-house counsel review. AI explains everything.
Secure fund transferMulti-factor verification. Fraud detection. $446M+ annual industry problem solved.
Recorded with county
Moving services connectedVetted movers, utility setup, relocation services
Closing complete. Ownership lifecycle begins.
Home Ownership Lifecycle
The platform doesn't end at closing. This is the retention moat — the digital twin and AI relationship persist for the life of ownership.
User-Controlled Proactiveness
Full Auto
"We've scheduled your HVAC tune-up for Thursday. Confirm or reschedule."
System books vendors, manages everything
Reminders Only
"Your HVAC filter is due for replacement. Want us to schedule it?"
System reminds, homeowner decides
Self-Managed
System tracks everything, homeowner handles it. Digital twin stays updated.
System logs, homeowner acts
Ongoing Services
Digital TwinUpdated with photos, receipts, notes — whatever is easy
Vendor NetworkTrusted, vetted vendors for any home need
Warranty TrackingExpiration alerts, claim facilitation
Insurance MgmtCompare, switch, annual review, claims
Property TaxProp 13 guidance, reassessment monitoring
Home ValueMonthly updates, equity monitoring, improvement ROI
When Ready to Sell Again
Homeowner ready to sell?
Digital twin is already completeYears of maintenance history, receipts, condition data
Quick re-scan for current conditionJust update what's changed since last scan
Listing ready in hours, not weeksEverything accelerated by existing data
Full seller flow beginsDigital twin transfers to new buyer at close
Rental Transition
Rent instead of sell?
Landlord mode activatedLease generation, tenant screening, rental income tracking
AI manages rental propertyMaintenance, tenant communication, compliance
Agent Experience (Trojan Horse)
How real estate agents use the platform — and how the system learns to eventually replace them.
Agent joins AdHome
License verified, onboardedAccess to tools & SaaS ($100-500/mo)
Agent gets powerful toolsTransaction mgmt, CRM, AI assistant, document tools, analytics
Agent uses AdHome for transactionsLists properties, manages clients, coordinates deals
System observes & learns from EVERYTHINGHow they negotiate, communicate, solve problems, make decisions
Patterns capturedDecision trees, communication scripts, objection handling, pricing strategies
AI handles more each monthRoutine tasks → communication → strategy → full transactions
AI capable enough?
Agent role transitionsAdvisory only, fewer transactions, or independent
The system IS the agent
The same pattern applies everywhere: Agents, loan officers, appraisers — build tools they love, learn from their expertise, progressively handle more, eventually operate independently. Every human interaction is a training opportunity.
Revenue Flows
How money flows through a single AdHome transaction — and how recurring revenue builds over time.
Single Transaction Revenue ($1.275M Bay Area Home)
$995
Seller Platform Fee
Buyers are FREE
$5,000
Embedded Lending
Origination + marketplace
$2,000
Title/Escrow
Vertical integration
$1,500
Vendor Services
5-15% take rate
$800
Insurance Referral
Year 1 commission
$500
Moving/Relocation
Referral fees
Total Revenue Per Transaction
~$10,695
vs. $63,750 traditional commission — consumers save 98%
Recurring Revenue Per Homeowner
Year 1: $1,200-2,000Insurance + premium sub + vendor marketplace
Years 2-7: $600-1,200/yearInsurance renewals + maintenance services + premium features
Resale (avg year 7-10): $10,695Full transaction revenue again — with accelerated listing from digital twin
Lifetime Value Per Homeowner (7 years)
~$20,915
Data & Learning Flywheel
How every interaction makes the platform smarter — and builds an unassailable data moat.
Transaction Happens
Data CapturedScan data, negotiation patterns, pricing outcomes, vendor performance, user behavior
Valuation Accuracy
Every sale price vs. AI estimate comparison calibrates the model. Interior condition data that no competitor has.
Negotiation Intelligence
Which strategies work for which property types, price ranges, and market conditions. Gets smarter every deal.
Matching Quality
What buyers actually buy vs. what they say they want. Implicit preferences outperform explicit criteria.
Vendor Performance
Real outcome data: which vendors deliver on time, on budget, with quality. Better recommendations over time.
Process Optimization
Where do deals get stuck? What causes delays? AI proactively prevents problems it's seen before.
Property Condition DB
Ground-truth, photo-verified condition data at scale. The most valuable asset — insurers, lenders, and investors pay premium.
All data feeds back into AI modelsContinuous improvement across all 17 systems
Better outcomes attract more usersMore users = more data = better outcomes = more users
Compounding data moatCompetitors can't replicate without processing the same volume
Data Monetization
Insurers
Photo-verified property condition data for underwriting and risk assessment. Reduces claims uncertainty.
Enterprise licensing: $50K-500K/year
Lenders
Interior condition data for collateral assessment. Supplements or replaces traditional appraisals.
Per-property API: $25-75/query
Institutional Investors
Hyperlocal market intelligence, pricing forecasts, inventory predictions, neighborhood-level analytics.
Enterprise: $100K-500K/year
Municipalities
Housing condition data for code enforcement, planning, and public policy. Anonymized and aggregated.
Government licensing: $25K-100K/year
At scale (100K+ properties): Data-as-a-Service alone could generate $25-50M ARR. This is the real moat — it compounds with every transaction and cannot be replicated without processing the same volume of guided scans.