# ClaimVision

> ClaimVision is an AI claims intelligence platform for US property carriers. It turns every page of a claim into structured data before your team opens the file, and lets them question that data with the evidence attached.

Last updated: 2026-10-09

ClaimVision is built by Decision Agency. The outcome it is bought for is claims cycle time: the reading is done before the file is opened, so the days a file waits come off the clock, inside the state prompt-payment and fair-claims-practices deadlines. The reading goes to the machine; the judgement stays with your people.

ClaimVision is not a replacement claims system. It is an intelligence layer that sits on top of the systems a carrier already owns, reads what is already there, and hands back judgement. The carrier's core system remains the system of record; the carrier's people keep the decision. ClaimVision started with the AI rather than adding it to an existing claims platform.

What it does, in business language: every page of the claim file read before the file is opened; a one-page view of a several-hundred-page claim; claims data your team can question in plain English ("are the solar panels covered at replacement cost?") that answers with the specific clause, the reasoning, and the dollar impact; coverage checked carefully and consistently on every claim; the fine print read together and explained in plain terms; figures held steady from the estimate to the letter; and a complete audit trail of every recommendation and every human override. The reading goes to the machine; the judgement stays with your people.

## The honest proof point

ClaimVision did a full day's reconciliation work on a real large-loss claim, done exactly, in under a minute. This is the only quantified claim ClaimVision makes today. Every finding ClaimVision returns is cited to its clause and page, and carriers can test it first on closed files they have already decided (the closed-file benchmark). It is built on real policies, inspection reports and claims, from first notice of loss to the decision letter. Broader results (leakage recovered, cycle time returned) are being measured on real, undecided claims alongside the human process; they will be published when measured, not before.

## Status

- Runs on Microsoft Azure and is built to SOC 2 principles. Decision Agency is working towards a SOC 2 Type 2 audit; ClaimVision is not yet SOC 2 certified
- A carrier's claim data never trains another carrier's models; each carrier's models learn from its own data
- ClaimVision is a product of Decision Agency. Legal entity: Decision Agency OÜ · Ahtri tn 12, 15551 Tallinn, Estonia · Registry code 17184257. Decision Crew is the separate consulting and integration firm.

## Pages

### Start here

- [ClaimVision: AI Claims Intelligence for US Property Carriers](https://claimvision.ai/): AI claims intelligence for US property carriers: every page of every claim read before the file is opened, each finding tied to its clause. No core swap.
- [What Is AI Claims Intelligence? Property Carrier Guide](https://claimvision.ai/ai-claims-intelligence): AI claims intelligence is software that reads every document in a claim file and returns findings with the evidence. How it differs from claims automation.

### Platform

- [Keep Guidewire, Duck Creek or Sapiens: Add an AI Layer](https://claimvision.ai/integrations): Add AI claims intelligence alongside Guidewire ClaimCenter, Duck Creek, Sapiens or in-house systems. ClaimVision reads the documents they hold. No migration.
- [AI Property Claims Review: Every Page Read](https://claimvision.ai/platform): ClaimVision reads the adjuster report, estimate, policy, endorsements and photos before your team opens the file, citing clause and page for each finding.
- [Ask Your Claims Data in Plain English](https://claimvision.ai/platform/ask): Ask questions of a property claim file in plain English. Every answer from ClaimVision carries its source clause, page and dollar impact, not a chatbot guess.
- [AI Claims Audit Trail for Regulators](https://claimvision.ai/platform/audit): An AI claims audit trail that keeps every finding, source, recommendation and human override, so carriers can answer regulators and defend each decision.
- [AI Coverage Checking for Property Claims](https://claimvision.ai/platform/coverage): AI coverage checking for property claims: the policy and every endorsement read together, each finding with its clause and dollar impact.
- [Security: Azure Hosting & Claims Data Handling](https://claimvision.ai/security): How ClaimVision protects claims data: Microsoft Azure in the US, isolated per customer, encrypted in transit and at rest, no cross-customer training.

### Solutions

- [AI for Property Adjusters: Training & Capacity](https://claimvision.ai/solutions/capability): Help new desk and field adjusters learn coverage faster: ClaimVision shows the policy clause behind every finding, so judgement grows on live claims.
- [Catastrophe Claims Surge: Hurricane & Hail AI](https://claimvision.ai/solutions/cat): Handle a hurricane, hail or wildfire claims surge with the same team: ClaimVision reads every file first so adjusters keep the same care at storm volume.
- [Reduce Property Claims Cycle Time](https://claimvision.ai/solutions/cycle-time): How property carriers cut claims cycle time inside state prompt-payment deadlines: remove the days a file waits to be read. With the NAIC deadline reference.
- [Claims Leakage in Property Insurance: Causes & Fixes](https://claimvision.ai/solutions/leakage): Claims leakage comes from missed exclusions, unapplied deductibles and figures that never reconcile. The five sources, each with a worked dollar example.

### Who it is for

- [AI Claims Tool for Property Claims Managers](https://claimvision.ai/for/claims-managers): For property claims managers: every overnight claim already read and sorted by what it needs, so desk adjusters spend the day deciding, not reading.
- [AI Claims Software for Claims Executives](https://claimvision.ai/for/executives): The P&L case for AI claims intelligence: shorter cycle time inside the prompt-payment clock, leakage and LAE that move with it, and no re-platforming.
- [AI Claims Platform for IT Leaders: Azure, No Migration](https://claimvision.ai/for/it-leaders): An Azure-hosted AI layer on your core claims system: no data migration, no new system of record, data isolated per carrier. What your security review needs.

### Comparisons (competitor facts from their public pages, checked 4 Oct 2026)

- [ClaimVision and Guidewire ClaimCenter: Add, Not Replace](https://claimvision.ai/compare/guidewire-claimcenter): Keep Guidewire ClaimCenter as your system of record and add AI that reads the whole property claim file. What stays in ClaimCenter and what the layer adds.
- [ClaimVision vs Shift Technology for Property Claims](https://claimvision.ai/compare/shift-technology): ClaimVision vs Shift Technology: a multi-line AI agent and fraud platform, or a property-only reader of the claim file that sits alongside your fraud stack.
- [ClaimVision vs Sprout.ai: Property Claims Compared](https://claimvision.ai/compare/sprout-ai): ClaimVision vs Sprout.ai for US property carriers: multi-line claims automation or a reader of the whole property file. Sourced side-by-side, checked Oct 2026.
- [ClaimVision vs Tractable: Auto Photos or the Whole Claim?](https://claimvision.ai/compare/tractable): ClaimVision vs Tractable: photo-based auto appraisal, or the whole property claim, with photos read alongside the policy, estimate and reports. Sourced.

### Resources

- [AI in Property Claims Ebook: From Hindsight to Foresight](https://claimvision.ai/resources/ebook): Free 26-page ebook for property insurers: the pressures on claims, where AI pays back and how to move from pilot to production. Every figure sourced.
- [ClaimVision FAQ: 100+ Questions Answered](https://claimvision.ai/resources/faq): Direct answers about ClaimVision: what it is, whether it replaces your claims system, data and security, cycle time, leakage, CAT surge and cost.
- [Property Claims Glossary: 36 Terms with Examples](https://claimvision.ai/resources/glossary): Plain definitions of US property claims terms (ACV, RCV, recoverable depreciation, proof of loss, wind/hail deductible and more), each with a dollar example.
- [AI Claims ROI Calculator for Property Carriers](https://claimvision.ai/resources/roi): Model AI claims ROI on your own assumptions: claim volume, review hours, adjuster cost, leakage rate and price. Worked example included. Not a quote.
- [Property Claim Walkthrough: Wind & Hail Example](https://claimvision.ai/resources/walkthrough): Follow one wind and hail claim on an HO-3 policy through ClaimVision in six screens, from upload to coverage findings to the decision letter.

### Guides and essays

- [Blog: Property Claims, AI and Adjusters](https://claimvision.ai/blog): Guides and essays on AI in property claims: unstructured documents, claims automation vs intelligence, CAT surge, coverage judgement calls and audit trails.
- [Property claim example, upload to letter](https://claimvision.ai/blog/a-property-claim-walked-through): A property claim example, start to finish: one fictitious wind and hail claim in Oklahoma, its seven coverage traps, and the calls its adjuster makes.
- [Claims Automation vs Claims Intelligence: Key Differences](https://claimvision.ai/blog/claims-intelligence-vs-claims-automation): Claims automation moves the claim; claims intelligence reads it. What each does, where automation stops, and which a property carrier needs first.
- [What a claim file should remember | ClaimVision blog](https://claimvision.ai/blog/file-memory): Most claim files keep the conclusion and lose the reasoning. Why that costs twice, in disputes and lost know-how, and why ClaimVision's record is append-only.
- [Ambiguous Policy Language: What AI Should Do](https://claimvision.ai/blog/judgment-calls): When policy wording reads two ways, AI should not decide silently. Present both readings with their dollar impact and let the adjuster rule.
- [Alternatives to Manual Review of Claims PDFs and Notes](https://claimvision.ai/blog/manual-claims-review-alternatives): Three alternatives to manual claims file review: more reviewers or outsourced desk review, extraction tools, or an AI layer that reads the whole file. Compared.
- [The most human we've ever been | ClaimVision blog](https://claimvision.ai/blog/most-human): The scarce resource in claims was never judgement or care. It was time, eaten by reading. What claims work looks like when the machine does the reading.
- [Why we publish exactly one number | ClaimVision blog](https://claimvision.ai/blog/one-number): AI vendor sites are walls of round statistics. Ours quotes one measured result and labels the rest as still being measured. Why claims auditors deserve that.
- [What 400,000 Retirements Cost a Claims Team](https://claimvision.ai/blog/retirement-wave): The claims workforce is retiring faster than it can be replaced. Why the real loss is judgment, not headcount, and what a claims team can do before it leaves.
- [Hurricane Claims Surge: You Can't Hire Your Way Out](https://claimvision.ai/blog/september): A hurricane multiplies the claims queue within days. Why CAT staffing plans answer volume but not variance, and what breaks first in surge week.
- [Unstructured Claims Documents: Insight Without a Core Swap](https://claimvision.ai/blog/unstructured-claims-documents): How property carriers turn unstructured claims documents (PDFs, estimates, notes, policies) into decisions without replacing Guidewire or Duck Creek.

### Property claims glossary (one page per term)

- [Actual cash value: definition](https://claimvision.ai/resources/glossary/actual-cash-value): Actual cash value (ACV) is what damaged property was worth at the time of loss, usually calculated as replacement cost minus depreciation. On a property.
- [Adjuster: definition and example](https://claimvision.ai/resources/glossary/adjuster): An adjuster is the professional who investigates a claim, establishes the facts, applies the policy and recommends or makes the payment decision. On a.
- [Bad faith: definition and example](https://claimvision.ai/resources/glossary/bad-faith): Bad faith is the legal exposure a carrier faces when it handles a claim unreasonably, for example through unreasonable delay, an inadequate investigation.
- [CAT event: definition and example](https://claimvision.ai/resources/glossary/cat-event): A CAT event is a catastrophe, such as a hurricane, hailstorm or wildfire, that produces a surge of claims across a region in a short window. It matters.
- [Claim file: definition and example](https://claimvision.ai/resources/glossary/claim-file): The claim file is everything the carrier holds on a claim: the first notice, the policy and endorsements, reports, estimates, photos, correspondence.
- [Claims leakage: definition and example](https://claimvision.ai/resources/glossary/claims-leakage): Claims leakage is the difference between what a carrier should have paid under the policy and what it actually paid, through overpayments, missed.
- [Coverage determination: definition and example](https://claimvision.ai/resources/glossary/coverage-determination): A coverage determination is the decision on whether, and to what extent, a policy responds to a loss, based on the policy wording and the facts of the.
- [Cycle time: definition and example](https://claimvision.ai/resources/glossary/cycle-time): Cycle time is the elapsed time from first notice of loss to resolution, usually measured to first payment or to closing. It is the measure families feel.
- [Declarations page: definition and example](https://claimvision.ai/resources/glossary/declarations-page): The declarations page is the summary page of a policy that lists the named insured, property address, policy period, coverage limits, deductibles and the.
- [Deductible: definition and example](https://claimvision.ai/resources/glossary/deductible): A deductible is the part of a covered loss the insured bears before the policy pays, set either as a flat dollar amount or as a percentage of the dwelling.
- [Depreciation: definition and example](https://claimvision.ai/resources/glossary/depreciation): Depreciation is the reduction in an item's value from age, wear and condition, applied line by line to turn replacement cost into actual cash value. It.
- [DP-3 policy: definition and example](https://claimvision.ai/resources/glossary/dp-3-policy): A DP-3 is a dwelling fire policy form commonly used for rental and other non-owner-occupied homes, covering the dwelling on an open-perils basis. Personal.
- [Endorsement: definition and example](https://claimvision.ai/resources/glossary/endorsement): An endorsement is a written change to an insurance policy that adds, removes or narrows coverage, or changes how a loss is settled. On a property claim.
- [Endorsement chain: definition and example](https://claimvision.ai/resources/glossary/endorsement-chain): An endorsement chain is the full set of endorsements on a policy, read together, to see which amendment controls a given loss and what is left of the base.
- [Exclusion: definition and example](https://claimvision.ai/resources/glossary/exclusion): An exclusion is policy language that removes coverage for specific causes, property or circumstances, such as flood, earth movement, wear and tear or.
- [First notice of loss: definition](https://claimvision.ai/resources/glossary/first-notice-of-loss): First notice of loss (FNOL) is the first report to the carrier that a loss has happened, usually from the family, their agent or a contractor. It opens.
- [Fraud signal: definition and example](https://claimvision.ai/resources/glossary/fraud-signal): A fraud signal is a pattern in a claim that warrants a closer look before payment, such as damage that does not match the reported cause, no recorded.
- [HO-3 policy: definition and example](https://claimvision.ai/resources/glossary/ho-3-policy): An HO-3 is the most common US homeowners policy form: it covers the dwelling and other structures on an open-perils basis, meaning anything not excluded.
- [Independent adjuster: definition](https://claimvision.ai/resources/glossary/independent-adjuster): An independent adjuster (IA) is a licensed adjuster who works for an independent adjusting firm or on contract, handling claims on a carrier's behalf.
- [Large loss: definition and example](https://claimvision.ai/resources/glossary/large-loss): A large loss is a claim big enough to need senior review or specialist handling, often six figures or more, with hundreds of estimate lines, several.
- [Loss run: definition and example](https://claimvision.ai/resources/glossary/loss-run): A loss run is a report of an insured's or a property's claims history, showing dates, causes of loss, amounts paid, reserves and status. Underwriters use.
- [Ordinance & law coverage: definition](https://claimvision.ai/resources/glossary/ordinance-and-law-coverage): Ordinance or law coverage pays the extra cost of repairing or rebuilding to meet current building codes, which the base dwelling coverage otherwise limits.
- [Peril: definition and example](https://claimvision.ai/resources/glossary/peril): A peril is a cause of loss, such as wind, hail, fire, lightning, water or theft. Coverage often turns on exactly which peril caused the damage, because.
- [Prompt-pay laws: definition and example](https://claimvision.ai/resources/glossary/prompt-pay-laws): Prompt-pay laws are state statutes and regulations that set deadlines for acknowledging, investigating, deciding and paying claims, with interest or.
- [Proof of loss: definition and example](https://claimvision.ai/resources/glossary/proof-of-loss): A proof of loss is a signed, sworn statement from the insured setting out the amount claimed and the facts behind it, such as the date, cause and value of.
- [Public adjuster: definition and example](https://claimvision.ai/resources/glossary/public-adjuster): A public adjuster is a licensed adjuster hired by the insured, not the carrier, to prepare and negotiate their claim, usually for a percentage of the.
- [Recoverable depreciation: definition](https://claimvision.ai/resources/glossary/recoverable-depreciation): Recoverable depreciation is the depreciation held back from the first actual cash value payment that the insured can collect once repairs or replacement.
- [Replacement cost value: definition](https://claimvision.ai/resources/glossary/replacement-cost-value): Replacement cost value (RCV) is the cost to repair or replace damaged property with materials of like kind and quality, with no deduction for.
- [Reserve: definition and example](https://claimvision.ai/resources/glossary/reserve): A reserve is the carrier's running estimate of what a claim will ultimately cost, set early and adjusted as the facts develop. Reserves are not payments.
- [Salvage: definition and example](https://claimvision.ai/resources/glossary/salvage): Salvage is the carrier's right to take damaged property it has paid for in full and recover its remaining value, usually by selling it. The amounts are.
- [Scope of loss: definition and example](https://claimvision.ai/resources/glossary/scope-of-loss): The scope of loss is the itemized description of what was damaged and what work is needed to repair it, the factual base under the estimate. It matters.
- [Statement of loss: definition and example](https://claimvision.ai/resources/glossary/statement-of-loss): A statement of loss is the document that sets out what the carrier proposes to pay: the gross loss by coverage, depreciation, the deductible, prior.
- [Straight-through processing: definition](https://claimvision.ai/resources/glossary/straight-through-processing): Straight-through processing (STP) is when a claim moves from report to payment without manual handling because every check has passed: coverage is clear.
- [Subrogation: definition and example](https://claimvision.ai/resources/glossary/subrogation): Subrogation is the carrier's right, after paying a claim, to recover that payment from the third party who caused the loss, such as a negligent contractor.
- [Supplement: definition and example](https://claimvision.ai/resources/glossary/supplement): A supplement is an addition to an approved estimate for damage or costs found after the first inspection, such as rotted decking discovered when a roof is.
- [Wind/hail deductible: definition and example](https://claimvision.ai/resources/glossary/wind-hail-deductible): A wind/hail deductible is a separate deductible for wind and hail losses, usually set as a percentage of the dwelling limit rather than a flat dollar.

### Company and policies

- [About ClaimVision and Decision Agency](https://claimvision.ai/about): Why ClaimVision exists: AI that does the reading so claims people keep the judgement. Who builds it, how we work and what we have proven so far.
- [Compare AI Claims Software for Property Carriers](https://claimvision.ai/compare): ClaimVision compared with Sprout.ai, Shift Technology, Tractable and Guidewire ClaimCenter: sourced, dated, and honest about where each one fits better.
- [Book a ClaimVision Demo: Bring a Closed Claim](https://claimvision.ai/contact): Book a 30-minute ClaimVision demo on a closed claim you already know, or start with the closed-file benchmark: decided files, no integration required.
- [Cookie policy](https://claimvision.ai/cookies): The ClaimVision website sets cookies only if you accept analytics: Google Analytics, by consent, with each cookie listed by name, purpose and lifetime.
- [AI Vantage: The Newsletter on AI in Property Claims](https://claimvision.ai/newsletter): AI Vantage is the free ClaimVision newsletter for property claims leaders: what changed, what it means for a claims operation, and what to do on Monday.
- [Privacy policy](https://claimvision.ai/privacy): How the ClaimVision website handles your data: what we collect, why, how long we keep it and your rights under GDPR and CCPA. Claim data is under contract.

## Machine-readable data

- [FAQ as JSON](https://claimvision.ai/api/faq.json): 103 questions and answers.
- [Glossary as JSON](https://claimvision.ai/api/glossary.json): 36 US property claims terms with definitions and worked examples.
- [Full text for language models](https://claimvision.ai/llms-full.txt): this file plus every FAQ answer and every glossary term.
- [Sitemap](https://claimvision.ai/sitemap.xml)

## Six questions, answered in full

### What is ClaimVision?

ClaimVision is an AI claims intelligence platform for US property carriers. It turns every page of a claim into structured data before your team opens the file, and lets them question that data with the evidence attached. Built by Decision Agency, it reads the adjuster's report, the estimate, the policy and all its endorsements, and the photos; checks coverage carefully on every claim; ties each finding to a specific clause with its dollar impact; and keeps a complete audit trail. It does not make the decision; it makes sure the person making the decision has everything in front of them.

### Does ClaimVision replace my claims system?

No. ClaimVision is an intelligence layer that sits on top of the systems a carrier already owns. Your core claims system remains the system of record; ClaimVision reads the documents it already holds and hands back judgement: findings, reasoning, and reconciled figures. Nothing is migrated, and your team keeps working where they work today. Carriers don't have a software shortage; they have an intelligence shortage.

### How does ClaimVision handle endorsements?

By reading them the way they were written to be read: together. Real policies are layers of amendments that change each other's meaning. ClaimVision works out what they add up to for each coverage question, and shows the reasoning: the operative clause, why it applies, and the settlement consequence in dollars. That reasoning trail is what lets a claims manager defend a decision, not just make it.

### What documents does ClaimVision read?

Every document in the claim file: the policy and its full endorsement stack, the declarations page, adjuster and inspection reports, repair estimates down to the last line item, statements of loss, proof of loss, photos with their captions and labels, and correspondence. If it's in the file, it has been read before the adjuster opens the claim.

### How accurate is it?

Here is the result we have proven: ClaimVision did a full day's reconciliation work on a real large-loss claim, done exactly, in under a minute. Every finding is cited to its clause and page, from first notice of loss to the decision letter. Broader accuracy metrics are being measured on real, undecided claims alongside the human process. We publish what we have measured, not what we hope; and when ClaimVision is not confident, it says so and puts the question in front of a person.

### Is it auditable?

Yes, completely. Every AI recommendation, every source it cited, every confidence level, and every human override is preserved. Nothing is overwritten. A regulator, reinsurer, or court can see what the system found, what it recommended, what the human decided, and why. Consistency plus documentation is also the strongest practical protection against bad-faith exposure.

## Contact

- Book a demo: https://claimvision.ai/contact. Bring a closed claim and interrogate ClaimVision on a file you already know the answer to.
- ClaimVision is a product of Decision Agency. Decision Crew is the separate consulting and integration firm.
- Legal entity: Decision Agency OÜ · Ahtri tn 12, 15551 Tallinn, Estonia · Registry code 17184257.
- LinkedIn: https://www.linkedin.com/company/claimvisionai