Your risk team runs and changes its credit policy

A policy change means a ticket to engineering or a vendor, then weeks of waiting. Decisimo lets the risk team edit and compare policy revisions before release.

Start free See the process

The credit decision, from application to policy change

  1. Application arrives

  2. Bureau and fraud

  3. Knock-outs

  4. Affordability

  5. Risk score

  6. Credit policy

  7. Compare and release

The whole credit policy as one decision flow

  • Sources togetherBureau and fraud data wait on the slower source
  • Debt to incomeAn editor formula includes the new monthly instalment, with no code
  • Policy tableKnock-outs come first, then affordability and score, with a reason on every row
Consumer lending decision flow: Bureau and fraud, knock-outs, instalment, debt to income, risk score and Credit policy
Consumer lending decision runs from Bureau and fraud through the Credit policy table.

See what a policy change does before release

  • RevisionsEvery save creates one, and two revisions or releases can be compared
  • Impact analysisIt compares decisions on the same applications by measures your team defines
  • New releaseYour systems keep the same URL. An endpoint can require sign-off.
Decision-mix charts compare baseline and comparison revisions for approve/decline results and loan amounts
Impact analysis compares a baseline revision of the Credit policy table with a comparison revision.

Read a decline back to the row that decided it

  • Scorecard breakdownSee each predictor's input, band and points, summed to the final score
  • Fired rowIts reason, table and revision identify the rule behind the decline
  • Decision recordInputs and outputs can go to the lender's Postgres for portal browsing
Calculation breakdown gives a final score of 660; the fired credit policy row records a decline for excessive commitments
One test run explains a score of 660, then the row that declined on debt to income above its limit.

Limits, offers and collections from the same parts

  • Shared building blocksA limit is a value the flow computes. The same parts serve each.
  • Trained modelsYour uploaded model runs in the flow, with your cut-offs around it
  • CasesA case can run the same flows when a person or document is needed

Collections decisioning

EU AI Act and GDPR for automated credit decisions

  • EU AI ActAnnex III duties for creditworthiness apply from 2 December 2027
  • GDPR Article 22It applies to an automated decline today
  • Recorded reasonsEach row gives the lender a plain reason for the decision

The two laws in full

What stays with the lender

  • Demo caseEvery screenshot value belongs to a demo case, run on demo applicants
  • Lender's policyDecisimo runs it on data the lender supplied
  • Your data sourcesThe lender chooses the bureau and fraud sources

Questions we get asked

What is a consumer lending decision engine?

A consumer lending decision engine runs a lender's rules and scorecards on an application to produce a credit decision. Decisimo runs the application flow from bureau and fraud data through the credit policy table.

Can I use this for mortgages and secured lending?

A mortgage's eligibility and price can be one policy. See mortgage underwriting.

Put your credit policy in a decision flow

Start free