AT&T Fraud Prevention Platform
My role
UX Design
UI Design
UX Research
UX Strategy
Project length
5 months
Project Overview
AT&T retail employees and fraud analysts rely on internal tools to investigate suspicious account activity before approving new services, account changes, or upgrades. These investigations require reviewing large amounts of customer data across multiple modules while identifying potential fraud signals under time pressure.
Our team redesigned critical investigation workflows to help employees quickly understand account history, identify risk indicators, and make informed decisions without jumping between disconnected systems.
The Challenge
Fraud analysts were required to review dozens of data points before determining whether an account should be approved, escalated, or investigated further.
Important information was often scattered throughout multiple modules, making it difficult to quickly answer questions like:
Who owns this account?
Is this customer new or long term?
Have there been recent account changes?
Are there other users associated with this account?
Are external fraud signals indicating elevated risk?
The experience increased cognitive load and slowed investigations.
Understanding Users
We worked closely with stakeholders and internal subject matter experts to understand how fraud investigations were performed.
Key goals included:
Reducing investigation time
Making risk indicators immediately visible
Surfacing the most important customer information first
Supporting consistent decision making across employees
Solution
We redesigned the investigation experience around how analysts naturally evaluate an account rather than how the data existed within backend systems.
Instead of forcing users to search through disconnected information, each module surfaced the most relevant information needed for decision making.
Applicant Identity Module
One of the primary modules focused on helping analysts quickly establish trust in an account.
The redesigned experience consolidated information including:
Account holder name
Individual or business account type
Account tenure
Authorized users
Billing address and so on..
High risk values were visually emphasized while trusted signals were surfaced to help analysts build confidence without reading every field individually.
Design Decisions
Prioritized information hierarchy
Frequently referenced information appears first while supporting information is progressively disclosed
Visual risk indicators
Color coded status badges and fraud scores make anomalies easier to identify at a glance.
Modular layout
Breaking information into focused sections reduced cognitive overload and allowed users to scan information more efficiently.
Historical context
Previous investigations and account history provide analysts with additional context before making decisions.
Impact
Although specific business metrics are confidential, the redesign helped create a more intuitive investigation workflow that supported:
Faster information retrieval
Lower cognitive load
Improved consistency across investigations
Better visibility into fraud signals
More efficient decision making for internal employees
