Insurance · the claims mailroom
Scan the claims mail once. Every page indexed to its claim.
Open the envelopes, stack the pages and scan. CapturePoint 6 finds where each document starts, knows a loss notice from a proof of loss from a repair estimate, reads the claim and policy numbers, and checks them against your open claims. Only the pages it is unsure of wait for a person.
Free trial starts the first time you open it. No form.
Two jobs, one station
Claims mail by claim number. Agency paper by policy number.
Set up each job from a folder of your own mail, then teach it your forms as you confirm them. The 2-minute tour on the welcome screen shows you the steps first, on sample accounts payable documents.
Setting up the job
Four decisions, made once, for every morning after.
Decision 1
How does a new document start?
Automatically is the usual choice for claims mail: CapturePoint 6 decides which pages belong together. Add barcode separator sheets if your mailroom already uses them, or use only separator sheets for files you never want it to judge.
Decision 2
Which values index each document?
Claim number, policy number, insured and loss date are on most claim documents. Add your own, such as the adjuster or line of business, and choose for each value how sure it must be before it goes without review.
Decision 3
What are numbers checked against?
Import a list, such as a daily export of open claims. A claim number read from the page can be checked against it, and a value the page lacks can be found in it, such as the claim number from the insured’s name. A name that fits two open claims waits for a person.
Decision 4
Where does each document go?
A folder per claim number with a searchable PDF and a data file for your claims system to pick up, or Content Central with every value in its fields. Choose the folders and file name per document type.
A worked example
Tuesday's first stack, as the index it becomes.
Seven documents for four claims, scanned in the order they came out of the envelopes. Five are filed without anyone touching them; two wait, each with its reason.
- First notice of loss
- NGM-2026-004187-01
- Dana Whitlock · HO 2210478-11
- Filed
- Repair estimate
- NGM-2026-004140-01
- Marcus Whitfield · PA 5961402-69
- Filed
- Sworn statement in proof of loss
- NGM-2026-004102-01
- Rosa Iglesias · HO 1987305-02
- Filed
- Police crash report
- Waiting: no claim number on the page
- Invoice
- NGM-2026-004140-01
- Waiting: line total does not match quantity times price
- Release of all claims
- NGM-2026-003996-01
- Owen Barclay · PA 4410992-05
- Filed
- Field adjuster report
- NGM-2026-004187-01
- Dana Whitlock · HO 2210478-11
- Filed
Fictional claims and insureds. Columns are the job's index fields; your data file carries every value read.
Why a page waits
Four reasons, in plain words, on the page itself.
The page has no claim number.
Crash reports carry a local report number, and medical bills an account number. Choose the claim once. With an open-claims list, the insured’s name can often find it for you.
A number is not on your list.
A transposed digit, or a claim opened after this morning’s export. The value is shown next to the page so you can correct it.
The line math does not add up.
On estimates and invoices it checks each line: quantity times unit price against the amount. A mismatch waits, so a payment is not figured on a bad total.
It has not seen this kind of document.
Confirm what it is and which values matter. After a few, it recognizes the next one by itself.
Confirm with Ctrl+Enter. Each confirmation teaches the job, and documents still waiting are checked again, so the ones that now pass clear by themselves. After steady use most mail needs no review at all.
After the scanner
The claim file, in the system your adjusters already use.
For each document: a searchable PDF, or PDF/A if you archive, a data file with every value, and a text file. Folders and file names are built from the values, for example one folder per claim number and a file named by document type and date of loss.
Sending to Content Central instead? Each document lands in your claims catalog with its values in matching fields. A workflow rule there can email the adjuster the moment a representation notice is filed.
Medical pages in injury claims
Your claims job can recognize the claimant's authorization for release of protected health information and itemized medical statements as document types of their own. That matters downstream: in Content Central, permissions by document type can keep them with the people handling the injury instead of the whole claims floor.
The rules a claim file answers toMailroom questions
Can CapturePoint 6 feed our claims system?
It exports each document to a folder as a searchable PDF (optionally PDF/A) with a data file holding every value it read, named and filed by claim number, for your claims system to pick up. It also sends documents with their values to Content Central, SharePoint or OneDrive (work or school accounts), Google Drive, Dropbox and Nucleus One.
Do claim documents leave our building to be read?
No. Pages are read on the Windows PC by the scanner. A graphics card makes it faster but is optional.
Can one station run both claims mail and agency paper?
Yes, as two jobs: claims mail indexed by claim number and agency or underwriting paper by policy number. Set each one up from a folder of your own paper, and scanning is unlimited at each station.
What about documents that arrive by email?
Save the attachments to a folder and import them into the same job: PDF, TIFF, JPEG, PNG, BMP and GIF are all read the same way as paper. If you would rather send documents in automatically and get the data back, Paige does that in the cloud, by webhook, SFTP or download.
Take the 2-minute tour, then run this morning's mail.
Start the tour from the welcome screen, then set up your claims job from a folder of your own claim mail.