Use caseInsurance mailroom

Inbound mail, every channel, one sort

Monday brings 4,000 items by paper, email and portal. Each needs a type, a case and a team, and the morning goes into reading envelopes.

Every channel lands in one sorting flowEach item reaches its case on its contentA person sorts what falls below the threshold

The capture resultType proposed: repair invoice

A plumber's invoice for a burst pipe in a kitchen. Type Repair invoice, confidence 97 percent, above the threshold for that type.

source · page 1 of the scan
Where the item goes

One point needs a person · the operator confirms or corrects

Actions on this inbound item
01

The starting point

In two paragraphs

A regional insurer receives about four thousand items on a Monday. The paper is scanned overnight, the emails sit in the claims mailbox, the portal uploads in a third queue. The hard part is the first reading: finding the policy number on page two, spotting a medical report inside an envelope, guessing which team owns a letter that names three contracts.

One flow now takes the first reading. Every channel feeds the same capture chain, the AI gives each document a type and a case on its content, and a threshold per type decides what goes straight through. The operators spend the morning on the items below the threshold, and decide each one.

Mailroom, Monday 08:30Capture
Three channels, one count. The night's mail is read before anyone logs in, and the items that stopped wait with their reason.
Inbound flow, MondayOperations engine
Ten o'clock. The three channels counted in one flow: 2,811 documents already with their teams, 296 waiting for a person.

Seen one at a time, the screens look familiar. Put them in order and the point shows: every channel is read the same way, each document reaches its case before the teams log in, and a person sorts what the AI cannot place. The walkthrough opens on the overnight batch and closes on the noon review of the corrections.

02

One flow for every channel

Before anyone reads a word, paper, email and portal uploads are gathered and split into documents.

Step 1Intake

The overnight scan batch is split into documents

The scanning provider dropped batch B-1145 in a shared folder overnight: 1,262 envelopes, each one a single file that may hold a letter, an invoice and a photo. The capture chain picks the batch up and splits each file document by document.

Batch B-1145Capture
One envelope often holds several documents for the same case.
Each document stands on its ownA letter and the invoice stapled behind it no longer share one file. Each can be typed, routed and found by itself.
Step 2Intake

Emails and portal uploads join the same chain

The claims mailbox holds 418 new emails and the customer portal has 301 uploads. Each email becomes an inbound item: the message is kept as a document, each attachment becomes its own. Portal uploads arrive through the API with the customer's account, so they already carry a customer number.

Paper, 1,262 envelopesEmail, 418 messagesPortal, 301 uploads
Email from j.porterViewer
The email opens in the browser like any scan, with its attachments listed.
Every item keeps its channel and its timeThe arrival time is what a response deadline counts from. It is recorded on the item, whatever door it came through.
Step 3AI check

Each document gets a type and a confidence

The AI classifies each of the 3,107 documents on the insurer's own scheme of 46 types: claim declaration, cancellation request, change of banking information, medical report, repair invoice, complaint. Each type comes with a confidence level.

The threshold is set per type. A complaint needs 90 percent, because a missed complaint starts a response deadline. Below its threshold, a document waits for a person.

Classification, batch B-1145Capture
The threshold belongs to the document type, and the administrator sets it.
The routine mail is typed before nineMost documents are clear, and they go straight through. What is left for a person is a short list with a reason on each line.
03

The case and the team

A typed document still needs an owner. The AI finds the case on the content, the process finds the team, and the lead sees what is left.

Step 4AI check

The case is found on the content of the letter

On the letter from James Porter, the AI reads policy HAB-55120-02 in the reference line and claim SI-26-04471 in the second paragraph. The repository confirms the claim is open on that policy. The letter is attached to it, whatever the file was called.

Letter D-30418Viewer
Each field read points to the line it comes from.
The mail reaches the case before the handler doesWhen the claims team logs in, the letter is already in the claim, with its type and its references.
Step 5Workflow

The routing rules send each item to its team

The claims process routes by document category, product and region. A home claim document about a house in Columbus lands with the home claims team there. A cancellation goes to customer service, a change of banking information to the back office.

Claim document, home: Home claims, ColumbusCancellation: Customer serviceMedical report: Health claims

Medical reports go to a team whose rights are restricted to that document type. Nobody else sees them in a queue.

Team queuesCase management
Each item lands in a team queue and in its case at the same time.
Each team starts with its own mailNobody forwards a letter to the right colleague. The route is a rule in the process, and the rule is the same for every channel.
Step 6Reporting

The mailroom lead sees what is left

Nadia, who leads the mailroom in Columbus, opens her home page at 8:30 a.m. One report counts the morning's documents by result. 296 wait for a person, and the possible complaints are at the top because of their deadline.

Mailroom, MondayCase management
The result of each document is a field like any other, so it can be counted and filtered.
The human work has a sizeNadia knows at 8:30 a.m. how much sorting is left and why. She plans the morning on a number.
04

The items that need a person

296 documents below the threshold. Nadia's operators clear them before eleven, in three moves.

Step 7Review

Nadia starts with the possible complaints

D-30414 is a letter about a five-week repair delay. It calls the situation unacceptable and asks for nothing specific. The AI proposed Complaint at 71 percent, under the 90 percent threshold, so it stopped.

Nadia reads it in the viewer and confirms the type. The letter goes to the complaints desk with its arrival time intact.

Item D-30414Capture
The operator confirms or corrects each value. Nothing leaves the queue without her sign-off.
Ask yourself: how many complaints sit in a general inbox for two days while their response deadline runs?
The deadline starts on the right dayA complaint is recognized the morning it arrives, even when it does not say its name.
Step 8Review

A choice between two open claims

D-30502 is a plumber's invoice for Rachel Lawson, who has two open claims: a water leak in March and a broken window in October. The AI proposes the water leak, because the invoice describes a pipe and dates the work March 18. It leaves the choice to Nadia.

Item D-30502Viewer
The proposal comes with the line it rests on. The operator picks the case.
The AI never picks between two casesIt says which one it would choose and why. Nadia confirms in one click, and the invoice joins the claim.
Step 9Settings

Corrections become examples for the next batch

At noon, Marcus, the capture administrator, reviews the morning's corrections. Eleven letters typed as Reminder were changed to Complaint by the operators. He adds three of them as examples to the complaint type, and adds a route for the pet insurance product launched in November.

Capture settingsAdmin console
Types improve with a few examples. Routes change in the process, with no redeployment.
Tuesday starts with Monday's lessonsThe operators' corrections feed the types. The administrator changes the settings, and the operators never touch them.
05

The outcomes

By 10:52 each of the 3,107 documents sits in one of three states, and each state sends it somewhere else.

Routed

Type and case are clear

The type is above its threshold and one case matches. The document is filed in its case with its fields.

Then: the team works it from its own queue.
To sort

One point needs a person

A type below threshold, or two cases that fit. The queue says which, with the line it rests on.

Then: the operator confirms or corrects, and the item moves on.
No case

Nothing on file matches

No policy or claim found for this customer. The AI lists what it read and where it looked.

Then: a person opens a new case, or sends the item back to the customer.
What the AI does not do

It opens no file on a guess. It picks no case between two candidates. It changes no setting. Across the nine steps on this page, every uncertain item is settled by an operator, with her name on the decision.

The AI reads, types and proposes. The thresholds and the routes are written by people, so the mailroom keeps control when Monday volumes double.

06

What it is worth

Taken from Monday's batch in Columbus. Your mail mix and your list of types will give other numbers.

1 queuefor every channel. Paper, email and portal are sorted with the same types and the same routes.
0 fieldskeyed in to attach a routed document. Policy, claim and customer are read from the document.
296 itemsleft for a person, each with its reason, out of 3,107 documents read that morning.

Where these numbers come from. They are counted on the scenario described on this page, with example data. They are not a customer result.

Three principles behind the figures

Zero wasted readingProductivityThe first reading is done before the teams log in. Operators read only what stopped.
Zero lost itemsProcess reliabilityEvery item keeps its channel, its time and its case. Nothing waits in a personal inbox.
Decision supportA person decidesEach proposal comes with its confidence and its source line. The uncertain ones stay with the operator.
07

For your teams

The mailroom, the capture administrator and IT each have a reason to want this, and a reason to doubt it.

Business operations

Mailroom operators and leads

Mornings go into opening and reading, and the hard items come last.

  • One queue for paper, email and portal.
  • Each stopped item says why it stopped.
  • Two candidate cases shown side by side, with the line that decides.
  • The size of the sorting work known at the start of the day.
Administration

Capture administrators

Every new product or form needs a sorting rule that nobody wrote down.

  • Document types on your own scheme.
  • A threshold per type, set where the risk is.
  • Types improved with a few confirmed examples.
  • Routes changed in the process, with no redeployment.
IT and architecture

CIOs and architects

Three channels, three tools, three ways of failing.

  • Portal uploads come in through the API.
  • Scans are picked up from the provider's shared folder.
  • Every AI call is logged with its model and its date.
  • A small model can run on premises, so no document leaves your network.

Start small

With one channel and ten document types. Take the scan batch, type it, and let the operators sort what stops. A few weeks of real mail will tell you whether it holds.

The other channels join the same chain afterwards, with the same types and the same routes.

See it live

Sort one of your own mail days

Bring an anonymized sample of a day's inbound mail and your list of document types. We will read it the way this page does, then sort what stops with your operators.

The screens on this page are drawn to show one step at a time, with example data.

Uxopian Software helps organizations put AI to work across the systems they already run, without giving up control of their data, their infrastructure or their critical processes.