Risk Monitoring System01 / 10

Manual monitoring became a business risk.

Internal tool for monitoring high-risk merchants.

Priority increased after missed cases and an audit fine.

Context
Automated monitoring for high-risk merchants
Starting point02 / 10

The BRD already contained UI decisions.

But the underlying workflow was not sufficiently confirmed.

BRD · page 2
Each monitoring rule was already represented as a separate tab.
All alertsExpected amountTrxns amountAverage ticketSuccess countDecline countRefunds amountNo alerts
Discovery03 / 10

I went back to the workflow.

2 interviews → insights → unresolved questions → job stories.

Discovery evidence
Interview insights table will be added here.

Waiting for the original Interview 1 — Insights file.

Prototype04 / 10

I prototyped the whole system in 7 days.

To validate the end-to-end workflow before detailed design.

End-to-end prototype
Architecture first. Detailed UI second.
Navigation05 / 10

Rule type was not the main unit of daily work.

My queue + statuses became the operational views. Group by rule remained optional.

Before
Rule-centric navigation
Expected amountTrxnsRefunds
After
My queue + operational statuses
NewInvestigationAwaiting replyClosed
Ownership06 / 10

One merchant case — one owner.

Taking one alert assigns all active alerts of that merchant.

Assignment rule
Assign one alert → take the active merchant case
Alert 01Alert 02Alert 03
Investigation context07 / 10

Analysts needed more than the current alert.

Active alerts · Closed alerts · Rule history

Alert details
Active alertsClosed alertsRule history
Rules08 / 10

Global defaults + Custom overrides.

A merchant inherits defaults until a specific rule is changed manually.

Rules model
Default → Custom only when changed
Default 35%Custom 40%
No alerts09 / 10

No alerts can also require review.

For a high-risk merchant, missing alerts may indicate incorrect rule settings.

No alerts workflow
Review merchants that remain silent.
Manage rulesCheck later
Delivery10 / 10

Final design → development.

Final ng-zorro screens were implemented.

Production outcome is unknown: launch was delayed by organisational and authorisation changes.

Delivery
Implementation-ready states and workflows