Use caseRetail banking

Completeness checks on your rules

Every home loan gets the same completeness review, from a PDF checklist that changes twice a year. Each processor reads it a little differently.

The administrator writes the rules as plain sentencesEvery verdict names the rule and the document behind itThe processor keeps the final word

The verdict on the application

One rule needs a person · the AI names it, with its document

Actions on this application
The assistantCompleteness check

Eleven rules pass. One does not: three recent pay stubs are needed for each borrower, and Anthony's August pay stub is not in the file.

source · 9 documents on the application
01

Why it drags today

In a nutshell

Columbus passes a file that Cincinnati sends back. Same bank, same checklist, the same PDF in the procedure folder, rewritten twice a year with an email from operations each time. Each processor applies the list from memory. The checklist exists. Nobody can say which version a given file was checked against.

Sophie Langford, who owns the procedure, writes the checklist once, as sentences. The assistant applies it to every application on arrival and writes a verdict with its reasons. She changes a sentence without a release, and the processor decides with the reasons in front of him.

Prompt libraryAI console
Sophie's prompt library. The completeness check is one business task, with its instructions, its model and where that model runs.
Home loan applicationsCase management
Matt's queue on Monday. The AI verdict sits in a column like any other field.

Rules in a console, verdicts in a queue: neither screen is new. The order is what earns the trust: the rules are written in sentences, tried on closed files, approved by compliance, then run on every new application with a reason for each result. The steps below start with Sophie's edit on November 3 and end with her month-end report.

02

Writing the rules

The procedure changed on November 1, and the administrator updates the checklist herself.

Step 1AI console

The checklist becomes sentences in the prompt editor

Sophie Langford runs credit operations for the bank's Ohio region. She opens the task Check completeness, home loan in the console. Its instructions hold eleven rules, each one a sentence. The new procedure wants pay stubs dated within 90 days, so she rewrites rule 4.

Prompt editor · Check completeness, home loanAI console
The instructions change without a release of the case application.
The procedure owner changes the ruleNo ticket to IT and no development cycle sit between the procedure and the files. The sentence Sophie writes is the sentence that runs.
Step 2AI console

One rule stops a file, and says why

Some rules end the review. At this bank, the home loan covers a primary residence or second home; a condo bought to rent out goes through the investment property loan instead. Sophie writes it as rule 12, with the verdict it returns and the sentence the processor will read.

Rules 1 to 11: Attention required if they failRule 12: Not eligible if it fails
Prompt editor · Rule 12AI console
The reason is written for the processor, who may repeat it to the client.
A refusal comes with its ruleThe AI lists what passes first, then names the rule that fails. Nobody reads a score and guesses the reason.
Step 3AI console

The model runs at the bank's site

Loan files hold income, tax and identity data. Sophie and the IT team pick a model hosted on the bank's own servers for this task, and set it in the task. The routing rule on data residency keeps every call for this case type on premises.

Task settings · ModelAI console
The model is chosen per task. Changing it needs no redeploy of the applications.
The data stays where the bank decidedCompliance does not have to approve a new place for loan files to go. The check runs where the files already live.
03

Trying it before it counts

The rules touch every application in the region, so they are tried on closed files and read by compliance first.

Step 4AI check

Three closed files, three verdicts to read

Sophie opens three applications the bank decided in October and runs Check compliance on each. Two verdicts match what the processors found at the time. The third flags a missing pay stub on a file the bank accepted.

Application PI-26-10437 · AI analysisAI console
The reasoning names the documents it read, so the administrator sees where the rule went wrong.
Put it to your best processors: how many of your rules are written down exactly as they apply them?
The rule is tested on files whose outcome is knownA wrong sentence shows up on a closed file, while nobody depends on it yet.
Step 5AI console

The rule a closed file proved wrong is rewritten

The borrower works for a hospital that issues two pay stubs a month, one for salary and one for allowances. Sophie adds one sentence to rule 4 and runs the check again on the same file. It now returns Complete.

Prompt editor · Rule 4AI console
The added sentence answers the case the first version missed.
The fix is one sentenceSophie edits the wording herself and checks the result on the file that failed. The case application is not touched.
Step 6Workflow

Compliance approves the rules before they go live

Sophie files the twelve rules as a document on the procedure case and launches its approval. Anne Doyle from compliance reads each rule next to the procedure section it cites, and approves. Only then does Sophie make the new instructions live.

Approval · Home loan completeness rulesCase management
The approval is a workflow task with its owner and its date.
The rules have a second readerThe author does not approve her own checklist. Compliance signs the text that will run on every file.
04

The checklist on a live application

On Monday a new application reaches a processor's queue, already checked.

Step 7AI check

The verdict is on the application before anyone opens it

Julie and Anthony Carver apply for $285,000 over 25 years for a condo in Columbus. The check runs as a step in the process and writes two fields on the case: a status and the reasoning. Here the status is Attention required: Anthony's August pay stub is not in the file.

Application PI-26-11208Case management
A click on the reason opens the pay stubs in the viewer, at the right page.
The gap is found at the doorMatt Russo does not discover the missing month while preparing the underwriting file. The verdict names it the morning the application arrives.
Step 8Case file

The processor decides what the warning means

Matt calls Anthony, who was on unpaid leave in August. He asks for the employer's letter, which arrives the same afternoon on the case. Matt records his decision in the case fields, with his reason, and sends the file to underwriting.

Application PI-26-11208 · DecisionCase management
The AI verdict stays in the history. The processor's decision sits beside it, with his name.
The warning is a question for a personThe rule did its job: it named the gap. Matt knows the client, and the decision is his, written down.
Step 9Reporting

The administrator sees which rule needs a person most

At the end of the month, Sophie opens her report of applications by AI verdict, then searches the files marked Attention required on rule 4. The processors' reasons repeat: unpaid leave, parental leave. She drafts a sentence on documented leave for the next approval.

Home page · ReportsCase management
The verdict is a reporting axis, and the processors' reasons can be read file by file.
The people who apply the rules improve themEach decision a processor writes down tells Sophie where the next sentence belongs. The checklist gets better from the files themselves.
05

The verdicts

The Carvers' application came back with one rule open. Every application gets one of three results, and the result decides what the processor does next.

Complete

Every rule passes

Identity, income, tax return, agreement, purpose. The reasoning shows each of the twelve rules as passed.

Next: the processor checks and sends the file to underwriting.
Attention required

One rule needs a person

A pay stub missing, a signature absent. The reasoning says which rule, on which document.

Next: the processor requests the document or records his own reason; the check runs again when a document arrives.
Not eligible

A rule stops the file, and you know which

The AI lists what passes first, then names the rule that fails and the product that fits.

Next: the processor reviews it with the client and points to the right product.
The limits we set

It refuses no loan. It sends nothing to underwriting. Across the nine steps on this page, every binding choice is made by a person: the administrator writes the rules, compliance approves them, the processor decides on each file.

The AI applies the sentences and writes its reasons. The control is in how the rules are written and approved, so it holds when volumes rise.

06

What it is worth

Counts from Sophie's November: one edited rule, one new one, three closed files. A longer checklist would call for more trial files.

12 ruleswritten as plain sentences by the person who owns the procedure.
0 releasesto change a rule. The instructions change in the console, the case application stays as it is.
3 closed filesused to try the rules before they ran on a single new application.

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

What makes these gains possible

Rules in sentencesOwnershipThe administrator writes and edits the checklist. Compliance approves it before it runs.
A reason per verdictProcess reliabilityEvery result names the rule and the document. The processor never decides on a bare label.
Decision supportA person decidesThe AI verdict stays in the history. The processor's decision sits beside it, with his name and reason.
07

Who uses it

This one needs an administrator, a processor and a compliance officer on board, and each of them sees a different part of it.

Administration

Credit operations administrators

When the procedure changes, you cannot tell who is applying the new version.

  • The checklist is a set of sentences you edit yourself.
  • A change needs no release of the case application.
  • You try the rules on closed files before they count.
  • A report shows which rule sends files to a person.
Business operations

Loan processors

The missing page turns up while you prepare the underwriting file, days after the application came in.

  • The verdict is on the application before you open it.
  • The reason names the rule and opens the document.
  • The request to the client can be drafted from the file.
  • Your decision and your reason stay beside the AI verdict.
Risk and compliance

Compliance officers

Approving rules you cannot read is approval in name only.

  • The rules are plain sentences, each one citing its procedure section.
  • The checklist goes live only after your approval.
  • Every analysis keeps its question, answer, model and time.
  • The model for loan files runs on the bank's own servers.

A good first step

Write the completeness review your processors do by hand as sentences, and run it on a handful of closed files before it touches a live one.

Your loan applications, documents and procedure stay put. The checklist runs inside the case you already use.

Let's talk

Write your own checklist with us

Bring your completeness procedure and a few anonymized files. We will write the rules with your administrator and run them on your closed files together.

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.