Skip to main content

The Claims Intelligence
Company.

Built entirely for the claimant side.

368M
FL property data points — indexed, enriched, and exclusive to the claimant side
FL Dept. of Revenue · 10,834,415 parcels · 34 fields · plus 100,412 carrier conduct records · 11 live integrations

100,412 carrier conduct records. 6-pillar intelligence platform. No competitor has this combination.

Built on one founder's 27x recovery.
Scaled to serve every claimant.

LossIntel Inc. is a Florida-based AI company that turns claims intelligence into outcomes — for consumers, licensed adjusters, and enterprise partners.

10.8M
FL Parcels Indexed
FL DOR cadastral dataset · 34 attributes each · 11 live API integrations
100,412
Carrier Conduct Records
FL DFS Civil Remedy Notices · 463 carriers · 2022–2026
27×
Founder Recovery Ratio
$23,266 first offer → $623,435 recovered · negotiation active

In 2024, three hurricanes struck Sarasota, Florida in 99 days — Debby in June, Helene in September, and Milton in October (Cat 5 at peak in the Gulf). Founder Stuart Nixdorff's property was in the direct path of all three. The carrier's first offer: $23,266.

He built the AI himself — document extraction, damage classification, settlement benchmarking against 100,412 carrier conduct records, statutory analysis, FEMA and SBA eligibility modeling. Total recovery to date: $623,435 across insurance, FEMA, and SBA. A 27× ratio on the first offer. The negotiation is still active.

That 27× gap — and the thousands of hours of documentation, calls, and appeals it required — is the entire product thesis. The information asymmetry between carriers and claimants is structural. LossIntel closes it.

LossIntel is pursuing FL Public Adjuster Firm Licensure to operate ClaimRestored as a direct licensed AI-native advocate — not a software layer on top of traditional PAs, but their AI-native successor. The model: AI does 95% of the work, a licensed PA supervisor validates and signs. One supervisor, 50–100× the throughput of a solo traditional PA.

The Proof Point That Became the Product
"Three hurricanes hit my house in 99 days. The carrier offered $23,266. I built the AI. We recovered $623,435. That's the product — every Florida storm survivor deserves the same outcome."
Stuart Nixdorff, Founder & CEO — LossIntel Inc.
Debby · Helene · Milton · 99 days · 27× ratio · Sarasota, FL

Yes. Ten select alpha users are live on the platform today with real underpaid Florida property claims in active review. 1,000 synthetic users validated the full pipeline — intake through evidence package — for QA and benchmarking accuracy. Mobile FNOL live: under 15 minutes from first notice of loss to a complete evidence-grade package. Public launch June 1, 2026 — the start of hurricane season — with the Founding 500 Florida storm survivor cohort. The YC batch is our first full hurricane season live.

Four structural forces converging simultaneously. The window is open now.

LossIntel is not a regulatory arbitrage play. It is a structural market inefficiency — and four external forces are converging to make this the right moment to close it.

Catalyst 1 · Regulatory

Florida Insurance Reform Creates Urgency

Post-Ian legislative reforms (SB 2A, HB 837) restructured the FL property claims landscape — tightening AOB restrictions, introducing a 60-day cure period under §624.155, and reshaping the public adjuster market. The window to build the definitive claimant-side intelligence platform is open now.

463
FL carriers in the CRN dataset · 100,412 verified conduct records
Catalyst 2 · Technology

AI Makes Expert-Level Analysis Scalable

Claude Sonnet 4.6 powers all 10 pipeline agents. ViT-L/14 damage classification, statutory state machine, and XGBoost settlement benchmarking all run at near-zero marginal cost. Intelligence that formerly cost $300–$500/hour is now accessible to every homeowner.

$0
Marginal cost per AI-generated gap analysis
Catalyst 3 · Climate & Peril

Every Peril. Every Season. The Asymmetry Never Changes.

Hurricanes, wildfires, tornadoes, water intrusion, hail, fire — the claim event changes, the carrier playbook doesn't. Low-ball. Delay. Deny. Whether it's Ian making landfall or a kitchen fire at 2am, policyholders face the same information gap against a carrier with 100× their data. Ian, Idalia, Debby, Helene, Milton gave Florida three hurricanes in 99 days in 2024 — but LossIntel's market isn't seasonal. It's every claim, every peril, every year.

18 yrs
Average FL homeowner tenure — hurricanes, fires, and water losses compound across a single lifecycle
Catalyst 4 · Mandate

HB 527 — Florida AI Mandate Creates Compliance Demand

Effective July 1, 2026, Florida HB 527 mandates that insurers using AI to deny or limit claims must provide human-reviewed explanations on request. Every LossIntel claim submission becomes an automatic HB 527 compliance trigger — putting carriers on notice and creating documented demand for appeal.

July 1, 2026
HB 527 effective date — FL AI denial review mandate goes live
Long-Term Architecture
Agent-to-Agent Endgame

In January 2026, Mighty completed the first documented AI-to-AI insurance claim settlements. Pactum AI has autonomously negotiated $140.5 million in commercial deals. LossIntel's trajectory: AI-powered evidence packaging today → agent-to-agent carrier negotiation by 2027–2029 → no-haggle settlement equilibrium thereafter. The mechanism is economic: contested claims cost carriers 3–4× standard claims. When both sides deploy AI, the rational outcome is faster, fairer settlement. LossIntel is building the claimant-side AI that makes that equilibrium inevitable.

Two brands. One intelligence layer. Full market coverage.

LossIntel Inc. deploys the same proprietary Claims Knowledge Graph and AI pipeline through two purpose-built brands — each optimized for a distinct segment and distribution model.

Brand 1 · Consumer D2C
ClaimRestored
claimrestored.com ↗
TurboTax for claim and benefit recovery.
AI-powered claims advocacy for homeowners, renters, and disaster survivors. Offers free claim prep before filing — then guided recovery for underpaid or disputed claims. Covers insurance, FEMA, and SBA disaster assistance in a single unified workflow.
Free Claim Prep — know your number before the adjuster arrives
AI policy parsing — finds missed entitlements, coverage gaps, and endorsement conflicts in the homeowner's own policy
ViT damage classification + settlement benchmarking against 100,412 CRN records
HB 527 compliance trigger — puts carriers on notice automatically
Any P&C claim: roof, water, fire, structural, theft
Direct consumer delivery in FL — PA partner channel in other states
$3B+
FL disputed claim value annually — primary beachhead market
Homeowner? Policyholder?
File your claim or check your settlement at ClaimRestored.com
File a Claim →
Parent Co · Data Infrastructure
LossIntel
lossintel.ai
The parent company and intelligence infrastructure behind ClaimRestored. The 10-agent AI pipeline, 100,412-record Civil Remedy Notice corpus, Claims Knowledge Graph, and BlindEye 6-pillar data stack all live here. LossIntel is pursuing FL Public Adjuster Firm Licensure — the endgame is not software for the incumbents we are replacing. It is becoming their AI-native successor: licensed, scalable, and running at a cost structure no traditional PA firm can match.
10-agent AI pipeline — Intake & Triage → Document Extraction → Damage Classification (ViT) → CKG Benchmark Lookup → Statutory Analysis → HB 527 Compliance → FEMA/SBA Eligibility → Settlement Gap Analysis → Evidence Package → Report Generation
Claims Knowledge Graph — 100,412 FL DFS Civil Remedy Notice records · 463 carriers · 2022–2026 · carrier dispute history, settlement ranges, geographic heat maps
BlindEye intelligence stack — 368M FL property data points across 6 pillars: cadastral (10.8M parcels · 34 attributes), permit history, flood zone, comparable sales, carrier conduct, weather event overlay
FL PA Firm Licensure — application in progress, target Q4 2026 / Q1 2027 · one licensed supervisor · 50–100× traditional PA throughput
Zero-friction FNOL — property pre-filled from 368M-record database at first notice of loss · voice intake, guided mobile walkthrough, and LiDAR/ARKit room scanning replace everything a traditional PA spends days gathering · zero marginal cost per claim
FNOL → Evidence Package
Under 15 minutes.
From first notice of loss to evidence-grade claim package. Voice intake, guided mobile FNOL, and LiDAR/ARKit room scanning replace everything a traditional public adjuster spends days gathering — at zero marginal cost per claim.
10
Autonomous AI agents — end-to-end pipeline
<15m
FNOL to evidence-grade demand package

From first notice of loss to evidence-grade claim package. Under 15 minutes.

Voice intake, guided mobile FNOL, and LiDAR/ARKit room scanning replace everything a traditional public adjuster spends days gathering — at zero marginal cost per claim.

📲
ClaimRestored
Policyholder Intake & FNOL Intelligence
Investor Deck — Confidential

From First Notice of Loss to Evidence-Grade Claim Package

Voice · Mobile · AI Vision · LiDAR/ARKit Room Intelligence · Automated Pricing — everything a public adjuster gathers, in minutes not weeks

🎙️
Voice Intake
Homeowner describes the loss naturally. AI structures it into a FNOL in real time.
Speech-to-textNamed entity extraction Date / peril parsingPolicy # lookup Multi-language
📲
Mobile FNOL App
Guided step-by-step intake. Pre-fills from BlindEye property data. No form-filling from scratch.
GPS address lockPolicy scan (OCR) Event date detectionPhoto + video capture Offline-capable
🏠
Visual Room Intelligence
LiDAR + ARKit scan maps every room, identifies all items, measures dimensions, prices replacements live.
ViT object detectionLiDAR / ARKit dims Live replacement costDamage classification Room-by-room log
Standard FNOL Data Points Captured
Policy & Policyholder
Policy numberCarrier name Coverage A/B/C/D limitsDeductible type Hurricane deductibleACC clause NFIP flood policyMortgagee info
Loss Event
Date of lossPeril type Hurricane name + CAT #Wind speed at property Storm surge indicatorPrior mitigation steps Contractor contactedPA involved
Property & Structure
Property addressYear built Sq footageConstruction type Roof type + yearStories Pool / outbuildingsElevation certificate
Damage Assessment
Roof damage %Water intrusion zones Interior rooms affectedStructural damage Mold / moisture readingsContents loss ALE / displacementCarrier offer received
Visual Room Intelligence — LiDAR + ARKit + AI Pricing Engine
📸 Room Scan Active — ViT object detection · LiDAR measurement · live pricing
🛌Sectional Sofa$3,200
📺75" OLED TV$2,800
🧪French Door Fridge$2,100
🪟Impact Window ×4$6,400
💡Lighting Fixtures$890
🪓Hardwood Floor$8,750
🚿Master Bath$14,200
🔌Appliance Package$4,600
847
Sq ft measured by LiDAR
43
Line items identified
$127K
Room replacement cost
Live Pricing Sources
Home Depot APILowe's API Amazon ProductRS Means Labor Xactimate CountersITEL Contents Marshall Swift
FNOL to Evidence Package — Automated Sequence
🎙️
Voice / Mobile
Loss Description
📸
LiDAR Room Scan +
Photo Capture
🔍
BlindEye
CKG Lookup
⚖️
Carrier D3
Risk Profile
📄
Evidence Package
+ Demand Letter
<15min
Full FNOL vs. days manually
43+
Line items auto-priced per room
80–150
Live attributes per property
7
Live pricing APIs cross-referenced
27×
Founder recovery vs carrier AI offer

Six data pillars. Ten AI agents. One outcome: the policyholder wins.

BlindEye is LossIntel's continuous insurance research engine and the proprietary moat. Six data pillars. Ten AI agents. 10.8M FL parcels with 34 attributes each. 100,412 carrier conduct records. 11 live API integrations. The carrier has had this infrastructure for a decade. Now the policyholder does too.

10.8M
FL parcels — 34 attributes each — 6-pillar intelligence stack, ingested and indexed
FL Dept. of Revenue cadastral dataset · 10,834,415 parcels · 34 attributes each · plus 100,412 CRN carrier records · 11 live API integrations
BlindEye — Claims Intelligence Platform
BlindEye — Pillar 1–6 · 368M Data Points · 100,412 CRN Records
Pillar 1 — Property Intelligence
Property Intelligence
FL DOR 10.8M parcels · 34 fields · 15+ live APIs per address
368M pts
Pillar 2 — Weather & Storm
Weather & Storm
NOAA 2010–2025 · Hurricane tracks · CAT serial numbers
7M pts
Pillar 3 — Benefits Model
Benefits Model
FEMA IA · SBA · Policy rules · ACC clauses · up to $687K path
4M pts
Pillar 4 — Insurance Law
Insurance Law
FL statutes · SB 2A · HB 837 · HB 527 · Bad-faith case corpus
2M pts
Pillar 5 — Carrier Intelligence
Carrier Intelligence
100,412 CRN records · FLOIR MIR 2022–2026 · Rate filings · 463 carriers
14M pts
Pillar 6 — BlindEye Playbook
BlindEye Playbook
Carrier D3 tactics · McKinsey docs · Xactimate counters · ITEL comps
105M+ pts
10-Agent AI Pipeline — Claude Sonnet 4.6 + XGBoost Settlement Benchmarking
01Intake & Triage
02Document Extraction
03Damage Classification (ViT)
04CKG Benchmark Lookup
05Statutory Analysis
06HB 527 Compliance Check
07FEMA / SBA Eligibility
08Settlement Gap Analysis
09Evidence Package Assembly
10Report Generation
🔒 Proprietary Data Moat — Not Replicable at Any Price
Underpayment Verdict
Probability score + recovery range
Demand Letter
Bad faith + FL § 624.155 citations
FEMA + SBA Path
Sequential filings up to $687K
Expert Network
Licensed PA · Attorney referral
Carrier Risk Profile
D3 pattern + negotiation strategy
368M
Property Data Points
100,412
CRN Records · 463 Carriers
13
Live Data Adapters
27×
Founder Recovery $623K vs $23K
10
AI Agents In Pipeline

The full machine: 6 data pillars → 10 AI agents → policyholder wins.

A compact view of the complete BlindEye intelligence flow for investors and technical partners — from raw data ingestion through the AI pipeline to policyholder-grade output.

👁
BlindEye
LossIntel Intelligence Platform
Investor Deck — Confidential

The Policyholder's Intelligence Advantage

368M data points · Symmetric AI against the carrier's Palantir stack

6 Data Pillars
Property Intelligence
10.8M FL DOR parcels · 34 fields · 15+ APIs
368M
Weather & Storm
NOAA 2010–2025 · Hurricane tracks · CAT serial #s
7M
Benefits Model
FEMA IA · SBA · Policy rules · ACC clauses
4M
Insurance Law
FL statutes · SB 2A · HB 837 · Case corpus
2M
Carrier Intelligence
100,412 CRN · MIR 2022–2026 · Rate filings
14M
BlindEye Playbook
D3 tactics · McKinsey docs · Xactimate counters
105M+
Ingest
AI Intelligence Engine
10-Agent AI Pipeline — Claude Sonnet 4.6 + XGBoost Settlement Benchmarking
01
Intake & Triage
02
Document Extraction
03
Damage Classification (ViT)
04
CKG Benchmark Lookup
05
Statutory Analysis
06
HB 527 Compliance Check
07
FEMA / SBA Eligibility
08
Settlement Gap Analysis
09
Evidence Package Assembly
10
Report Generation
🧠 ViT Vision Model
📊 XGBoost Benchmarking
🖷️ CKG Claims Graph
Redis · Celery
🐘 PostgreSQL
🔒 Proprietary Data Moat — Not Replicable
Output
Policyholder Value
Underpayment Verdict
Probability + recovery range
Demand Letter
Bad faith + statutory citations
FEMA + SBA Path
Sequential filings up to $687K
Expert Network
Licensed PA · Attorney referral
Carrier Risk Profile
D3 pattern + negotiation strategy
368M
Data Points Ingested
10.8M
FL Parcels Mapped
100,412
CRN Filings · 463 Carriers
27×
Founder Recovery $623K vs $23K
10
AI Agents In Pipeline

The Claims Knowledge Graph. 100,412 carrier conduct records, structured for AI inference.

The CKG is not a database — it's the ground truth layer. Every claim runs through AI policy parsing to find what the carrier owes, ViT damage classification to document what happened, and XGBoost settlement benchmarking to prove the gap.

Claims Knowledge Graph (CKG) — 100,412 FL CRN Records · 463 Carriers · 2022–2026
Florida's Department of Financial Services Civil Remedy Notice database, fully structured and enriched: carrier, claim type, dispute outcome, geography, settlement range, and regulatory action history. No public adjuster firm, no insuretech, and no carrier-side platform has this data in a queryable, ML-ready format.
Layer 1 — Data
Claims Knowledge Graph
CKG
100,412 FL DFS Civil Remedy Notice records across 463 carriers, 2022–2026. Carrier-level dispute history, claim type distribution, settlement ranges, geographic heat maps, and outcome benchmarks. The ground truth dataset for property claim intelligence in Florida.
Proprietary — no commercial equivalent exists
Layer 2 — AI Vision
Damage Classification
ViT-L/14
Vision Transformer model (ViT-L/14) classifies property damage from photos — roof, structural, water, mold, fire — and maps classifications to CKG benchmarks. Replaces what a licensed adjuster takes hours to assess.
Near-zero marginal cost per analysis
Layer 3 — Statutory
Regulatory State Machine
FL § 624.155
Automated bad-faith statutory framework. Tracks FL § 624.155 Civil Remedy Notice requirements, HB 527 AI disclosure triggers, 60-day cure period logic, and carrier-specific compliance history. Creates documented demand for appeal at every submission.
HB 527 effective July 1, 2026
10-Agent AI Pipeline — Claude Sonnet 4.6 + XGBoost Settlement Benchmarking
01 Intake & Triage
02 Document Extraction
03 Damage Classification (ViT)
04 CKG Benchmark Lookup
05 Statutory Analysis
06 HB 527 Compliance Check
07 FEMA / SBA Eligibility
08 Settlement Gap Analysis
09 Evidence Package Assembly
10 Report Generation
Competitive Position
The Empty Quadrant — No AI-Native Competitor on the Claimant Side

Verisk, CoreLogic, and Tractable all serve carriers. Traditional PAs and plaintiff attorneys are low-tech. LossIntel occupies the only unoccupied position in the market — high-tech, AI-native claimant advocacy.

✓ LossIntel — High Tech / Claimant Side Verisk — High Tech / Carrier Side Trad. PAs — Low Tech / Claimant Side
Data Flywheel
Every Resolved Claim Makes the Platform Harder to Replicate

Every resolved claim enriches the CKG. Carrier settlement patterns, underpayment ratios by zip code, and regulatory timelines create a compounding data advantage. A competitor starting today faces a dataset gap that compounds with every claim LossIntel closes — not just today's 100,412 records, but every record added going forward.

Policy Intelligence
AI Policy Parser & Gap Analysis
Claude Sonnet 4.6 parses the homeowner's own policy to identify missed entitlements, coverage gaps, and endorsement conflicts the carrier won't volunteer. Automated demand letter and FL DFS complaint generation included. Trained on 100,412 Civil Remedy Notice records.
policy-parser/gap-analysis-v2
Computer Vision
ViT-L/14 Damage Classification
Vision Transformer trained on FL catastrophic loss imagery. Classifies damage type, severity, and repair scope from photos — maps directly to CKG benchmark ranges and contractor cost data. >91% classification accuracy.
openai/vit-l-14 (fine-tuned)
ML Benchmarking
XGBoost Settlement Model
Trained on 100,412 CRN records. Predicts settlement ranges by claim type, carrier, geography, and damage severity. Outputs confidence intervals and carrier-specific dispute tendency scores.
xgboost/settlement-benchmark-v2
Statutory Engine
FL Regulatory State Machine
Tracks § 624.155 bad-faith requirements, HB 527 AI disclosure triggers, AOB restrictions, and carrier-specific compliance history. Automatically generates CRN filing triggers on every submission.
fl-regulatory-engine/v3

$1.06T U.S. P&C market. Florida beachhead: 150K–200K claims/year, 30K+ disputed, $3B+ dispute value.

42% of homeowner claims were closed with zero payment in 2024 — up from 25.7% in 2004. Every one of those claimants is a potential LossIntel user. The market compounds with every hurricane season.

The Carrier Playbook
Defend. Delay. Deny. — What used to be a secret backroom training manual is now a deeply embedded AI algorithm.
Insurance carriers manage $125B+ in catastrophic liability after a major storm year. Their playbook is optimized to exhaust the policyholder: drag the process out, minimize the initial offer, and deny where possible. That system was always asymmetric — but now it runs at machine speed. Carriers are embedding Defend-Delay-Deny logic into AI-driven claims adjudication tools with no visibility, no training manual to leak, and no human review. The policyholder on the other side has none of that. LossIntel was built to be the counter-intelligence layer — the first AI system built entirely for the claimant side.
$1.06T
Total Addressable Market
U.S. P&C premiums — AM Best / S&P Global 2024
$3B+
FL Annual Dispute Value
30K+ disputed claims — FL beachhead serviceable market
42%
Zero-Payment Claims (2024)
Homeowner claims closed with $0 — Weiss Ratings 2024
TAM
$1.06T
U.S. P&C premiums — total addressable
SAM
$3B+/yr
FL disputed claims — serviceable addressable market
SOM
$5M+
Year 1 FL target — claim value processed

Sources: U.S. P&C premiums (AM Best / S&P Global, 2024). FL claims volume (FL DFS Annual Reports, 2022–2025). Zero-payment claim rate (Weiss Ratings, 2024). Dispute value derived from FL DFS complaint data and NAIC market conduct reports. Year 1 SOM based on SW Florida beachhead: Lee, Collier, Sarasota, Broward counties.

Seed Round Q2 2026. Initiating institutional conversations.

LossIntel Inc. is raising a $2.5M seed round at a $10M pre-money valuation. Y Combinator Summer 2026 applicant. Gustaf Alstromer's thesis — "sell the work, not software" — describes our architecture exactly. The YC post-loss claims lane is open. LossIntel is in it.

Round
Seed
Q2 2026 · Initiating conversations with select institutional and strategic partners
Year 3 Revenue Projection
$5.52M
Year 3 revenue projection (modeled, illustrative) — see data room for full assumptions
Beachhead Market
$3B+/yr
FL beachhead — highest-volume property claims market in the US · HB 527 effective July 1, 2026
Regulatory Catalyst
July 1, 2026
FL HB 527 effective date — AI denial review mandate creates structural compliance demand
$280
Blended CAC
Paid search + content + referral (modeled, illustrative)
$3,100
Customer LTV
1.1× claims/customer + referral value (modeled, illustrative)
11.1x
LTV / CAC Ratio
>3x investable · >10x exceptional
82%
Gross Margin
AI-driven analysis at near-zero marginal cost (modeled, illustrative)
Request the Data Room
Full financials, market model, 6-tier revenue architecture, technology stack, and cap table available to qualified investors under NDA.

Six novels. Six disasters. One algorithm.

A moat no purely-technical competitor can replicate — and a head start no one can buy.

Blind Eye — A Thriller by Stuart Nixdorff
blindeye.ai ↗
Cultural Acquisition Layer — BlindEye Fiction Series

BlindEye publishes state-and-disaster-specific fiction novels as cultural acquisition and stakeholder pressure infrastructure. Each book is anchored on a real catastrophic event, a real insurer pattern, and a real claims outcome. The first novel — Blind Eye — is set in Florida during the 2024 hurricane sequence. Stuart Nixdorff, author.

Six novels in three years. Hurricane Milton (FL) · Camp Fire & Eaton Fire (CA) · Tornado Alley · Hurricane Harvey (TX) · Hurricane Ida (Northeast) · Helene / Idalia (FL Panhandle). Owned IP for LossIntel Inc. Published under the BlindEye brand.

Each novel functions simultaneously as customer acquisition, regulatory pressure, and cultural moat. A reader who finishes Blind Eye understands carrier bad faith, public adjuster leverage, and their own rights as a policyholder — and knows exactly where to go. No purely-product competitor can replicate this on the timeline we already have.

6
Novels · 3 years
6 states · 6 perils
Channel function
Acquisition · Pressure · Moat
IP
Owned by LossIntel
Inimitable on our timeline

Investors and partners.

LossIntel Inc. is actively seeking institutional investors, strategic partners, and press or regulatory engagements.