Credit scoring in plain language
The story operators tell
- You ask the system to score a person or application context the bank allows on screen.
- AIS reads the approved data fields (income, bureau, behavior, product, etc.-depends on integration).
- Models and rules approved by your bank produce a score or band plus short reasons where explainability is on.
- Decision PRO shows the outcome on Credit scoring pages and often links to history for audit.
What “good” looks like
- Stable identifiers (customer ID, application ID) on every row in history.
- Timestamps and version hints when your bank requires them for model governance.
What can go wrong (operator view)
- Missing data → incomplete score or “cannot score” message-fix data at source, do not invent values.
- Slow response → wait; if timeout, use your support path with time of event and user id.
Where to practice clicks
Use Decision PRO Credit decisioning guides and expand them with USER_TRAINING_CREDIT_SCORING.md in the monorepo for labs.
Next: Risk, policy, and models