Claimory AI already knows your claims.
Claimory AI is the assistant built into your collision repair workspace. Ask it about any claim, any car in the shop or any carrier you are fighting, and it answers from your own claim numbers, supplement history and photos, then proposes the next move. There are 63 moves it can carry out, and not one of them touches your workspace until you approve it.
A capture from a working shop workspace, not a rendering. The six numbers on it are the six notes around it.
- 1
Its own place in the navigation
Claimory AI sits in the main navigation next to Claims and Cash Jobs, and opens as its own page with every past conversation down the left.
- 2
Every conversation keeps its scope
Each thread carries a label under its name: Workspace, This claim, Claims queue or This cash job. Reopen one a week later and it still knows which claim it was about.
- 3
Your claim numbers, your carriers
The answer names the claims actually sitting in your shop, the carrier on each one, and how long each has been waiting. It is reading your workspace, not collision repair in general.
- 4
The next move comes with the answer
Under each answer it proposes what to do next: draft the adjuster email, create the follow-up task, pull the list. One press runs it, and anything that changes your workspace waits for that press.
- 5
Attach the evidence
Photos, carrier PDFs, documents and audio go straight into the conversation, and the chip on each file tells you whether it was actually read.
- 6
Every answer stays searchable
Search chats sits above the conversation list. The answer Claimory AI gave you three weeks ago about a claim is still there, still carrying the scope label of the claim it was about.
- What it does
- Claimory AI reads the claim you are standing in: the estimate, the photos, the carrier PDF and the supplement history. It catches the unbilled scan and the softened labor hour, drafts the rebuttal to a denial, answers in your own claim numbers and carriers, runs a full Claim Audit on an estimate, and many more.
- What it costs
- Claimory AI is included from Professional up, and chat is unlimited on Professional, Elite and Enterprise. Starter is the bare-bones tier and includes no AI at all.
- What it can change
- Every action that touches your workspace waits for you to approve it first. Web search is the single exception, because it reads the internet and writes nothing to your shop.
Claim Audit reads the whole claim at once
Chat answers the question you thought to ask. Claim Audit is the pass you run when you do not know what you missed: every photo, every line item and every procedure on one claim, read together, returned as a list of findings rather than a paragraph.
It finds the money the sheet left off
Missed labor operations, blends and prep work. The dollar gap shows beside each finding, so a finding is a number and not an opinion.
It names the lines a carrier will fight
Where the carrier will push back lists each line the carrier is likely to fight, with what they will say and your comeback, so the argument for the line is ready before you call the adjuster.
It writes the supplement note for you
Every finding carries a paste-ready note in your shop's voice, so the finding reaches the estimate in one copy.
Every finding carries approval odds
A 0 to 100 prediction that this carrier approves this line as written, with one sentence saying why, so you can push the lines worth pushing and drop the rest. The report multiplies the odds against the amount to show the dollars at risk on each finding.
Nothing is applied for you
The audit hands back a list. You take the lines you agree with and ignore the rest. It never edits the estimate on its own.
It runs on credits
Thirty audits a month on Professional and a hundred on Elite, unlimited on Enterprise, and three during the trial.
Eleven brains. One voice.
Claimory AI is not one general assistant told to act like an estimator. It is eleven specialist instruction sets, and the master loads only the ones your question actually needs. You never pick one, you never see the handoff, and one answer comes back.
Master
The voice. Decides who answers what.
Every question goes to the master first. The master decides which specialists to call, gathers what they send back, and returns one answer in your shop's voice. The other ten brains never speak to you directly. The master does the talking.
Operations
The hands. Turns intent into action.
Whenever the master decides something needs to happen in the workspace, operations is the brain that does it. Tasks, supplement drafts, emails, calendar events, tow logs, parts orders. Every action surfaces as an approval card you tap before it runs.
Estimator
The pro. Reads photos like a 20-year vet.
Twenty years of body shop estimating distilled into one specialist. Knows the manufacturer position statements, the labor time references, and where the missed money usually hides. Spots the unbilled scan, the missing blend, the underbilled R&I, and many more.
Advocate
The responder. Drafts your carrier replies.
When a carrier denies a line, the advocate drafts the response. Knows each major carrier's denial patterns and which citations actually move them. Drafts in your shop's voice, professional and factual, fact then citation then dollar impact.
Researcher
The librarian. Grounds every claim.
Pulls from a verified knowledge base before saying anything that needs to be backed up. Cites the source. Says so when no source is on file instead of inventing one. The reason the rest of the brains stay honest.
Vision
The eyes. Sees what humans miss.
Looks at the photos on the claim and reports what is in them: panel, damage type, severity, repair operation implied. Flags structural concerns up front because the dispatch is different. If a photo is unusable it says so plainly.
Communications
The host. Talks to customers and vendors.
Customer pickup texts, status updates, vendor purchase orders, internal team notes. Different tone than the advocate, since these go to your own customers and team. Plain language, time anchored, never invents a fact.
Compliance
The shield. State rules and DOI bulletins.
State-by-state regulatory specialist. Labor rate caps, storage rate caps, total-loss notification windows, right-to-appraisal triggers, anti-steering, aftermarket parts consent. Hedges hard when the rule is not in the verified knowledge base.
Strategist
The owner brain. Margin, mix, capacity.
When the question goes past a single claim, the strategist takes it. Margin trends, carrier scorecards, capacity, when to hire. Lead with one number, follow with one move, never recommend a big change off a single month of data.
Pitch
The intro. Loads only when you ask.
The brain that answers the existential question, what does Claimory actually do. Loaded only when someone asks, so the marketing copy never bloats every casual chat turn. Picks two or three concrete levers that match the asker's role and pain.
Workspace
The map. Knows every tab in your shop.
Knows every module in the workspace and how to read it: claims, cash jobs, tasks, calendar, email, SMS, inventory, loaners, tow, financials. The brain that orients the master when a question is about your shop's state, not a single claim.
Ask Claimory AI any of these, in your own words, and many more are being implemented. It comes back with the operation proposed as a card. Open a row to see exactly what would happen to your workspace if you approved it.
What you can ask Claimory AI to do in your shop
Ask Claimory AI any of these, and many more being implemented
Every command above changes something in your workspace, so every command above waits for you. Web search is the one exception, because it reads the internet and writes nothing to your shop.
It already knows what you are looking at
You do not paste a claim number to start. Claimory AI resolves the scope from the screen you are on, so "what is this one missing" means the claim in front of you. Thirteen scopes are recognized.
- claim
- financials
- tasks
- calendar
- sms
- reports
- inventory
- loaners
- towing
- settings
- dashboard
- workspace
And it reads what you hand it
Attach the evidence rather than describing it. The chip on each attachment reports whether the file was actually read, so you are never guessing whether the AI saw what you sent.
Photos. Teardown and damage shots, analyzed for what is missing
Carrier PDFs. Parsed into populated fields rather than retyped
Documents. Text extracted so the model reasons over the contents
Audio. The voicemail you would otherwise transcribe by hand
It opens on top of the screen you are already on
Claimory AI is not a page you navigate to and lose your place. The drawer is mounted once for the whole workspace, so it opens over the claims list, a claim record, the task board or the calendar without taking you off it, and it opens already holding what that screen is about.
One drawer, every screen
It is mounted once for the whole workspace rather than per page, so the same conversation is a keystroke away from the claims list, a claim record, the task board and the calendar.
It opens already knowing what is late
Before you type, it runs the shop's own counts and shows up to five: active claims untouched for seven days or more, overdue tasks, overdue loaners, unread carrier email, and supplements past your follow-up window.
Your follow-up window, not ours
The supplement card uses the number set in your workspace settings and falls back to seven days only if you never set one, so a shop that chases at three days is not told it is fine at six.
That first screen writes nothing
Every card is read-only. Tapping one fills the composer with the question and waits for you. Anything that changes a record still arrives as an action card you approve.
The useful part is what it will not do
An assistant that can carry out real operations on a live shop is only worth having if you can see exactly where it stops. Here is every brake, including the one exception.
Approval is a real gate, not a setting
Every action that changes your data arrives as a card with Cancel and Approve and Run. Until you press Approve, nothing is written. Cancel leaves no trace on the claim.
The one thing it does without asking
Looking something up on the public web. It reads, it never writes, and it changes nothing in your workspace. We would rather name the exception than claim an absolute that is not true.
Your data is not training data
Prompts carry claim context only, and your workspace data is never used to train provider models. Claimory AI answers your question and forgets it.
It can be disabled entirely
Disable AI for the workspace and the rest of Claimory keeps working exactly as it did. Claims, cash jobs, the portal, tasks and email do not depend on it.
Regulation answers cite their source
State regulation questions are answered from a sourced reference corpus, and the answer names the document it came from. If the answer is not in context, it says so instead of inventing one.
Nothing is auto applied
A Claim Audit returns a suggestion list. Missed blend, softened hours, overlooked R/I. You apply what you agree with and ignore the rest.
Every brake above is in the version you get on day one of the trial, and the trial includes Claimory AI.
No credit card. Works with CCC ONE and Mitchell.
Everything Claimory AI does
- Eleven specialist brains, routed per question, with more than one able to activate on a single turn
- 63 named actions Claimory AI can carry out against a collision repair workspace
- Every action that changes workspace data waits for explicit operator approval
- Reads the screen the operator is on across 13 context scopes
- Opens as a drawer over the screen the operator is already on rather than a separate page
- Shows up to five proactive cards before the first question: claims untouched for seven days or more, overdue tasks, overdue loaners, unread carrier email and supplements past the shop's own follow-up window
- Photo, PDF, document and audio attachments, with a read receipt on each file
- Claim Audit that flags missed line items, missing blend and softened labor hours with cited evidence
- Adjuster and customer message drafts written in the shop's own voice
- State regulation answers that cite the source document
- Carrier PDF parsing into populated claim fields
- Unlimited Claimory AI chat on Professional, Elite and Enterprise
- Approval odds on every Claim Audit finding: a 0 to 100 prediction that this carrier approves this line as written, with the reason for the number
- Dollars at risk on a finding, computed from its amount and its approval odds
- Provenance lock: a finding with no photo id and no estimate line number is dropped before display
- Teardown requests and an ADAS calibration matrix returned alongside the findings
The questions shops actually ask first
Ten questions about what Claimory AI reads, what it is allowed to change and what it costs, answered with the limits left in.
Put it on your own claims
Bring one collision claim you think is finished and ask Claimory AI what the estimate missed. Nothing in your workspace changes until you approve the action it proposes.
No credit card. Works with CCC ONE and Mitchell.