Designed a transactional fraud tool into a strategic dashboard that doubled resolution speed and motivated teams through gamified UX
Designed a transactional fraud tool into a strategic dashboard that doubled resolution speed and motivated teams through gamified UX
Designed a transactional fraud tool into a strategic dashboard that doubled resolution speed and motivated teams through gamified UX















On Track
On Track
On Track
On Track
On Track
On Track
[Context]
A credit card fraud detection company ("Phoenix") based in California partnered with SJSU to transform their fraud detection system using real transaction data and frontline team feedback. I led the design of the manager dashboard while coordinating closely with a teammate who designed the complementary analyst interface. We maintained consistent UX patterns so both tools worked seamlessly together.
[Type]
Industry-Academia
[My Role]
Lead Product Designer
[Timeline]
6 months
[Team]
Co-designer, Project Advisor
[Context]
A credit card fraud detection company ("Phoenix") based in California partnered with SJSU to transform their fraud detection system using real transaction data and frontline team feedback. I led the design of the manager dashboard while coordinating closely with a teammate who designed the complementary analyst interface. We maintained consistent UX patterns so both tools worked seamlessly together.
[Type]
Industry-Academia
[My Role]
Lead Interaction Designer
[Timeline]
Jan - July 25
[Team]
Co-designer, Project Advisor
TL;DR



[Problem]
[Problem]
Analysts struggled with high-volume manual case reviews, while managers operated without clear KPIs or fraud trend visibility—resulting in inefficiencies and low team morale.
Analysts struggled with high-volume manual case reviews, while managers operated without clear KPIs or fraud trend visibility—resulting in inefficiencies and low team morale.



[Outcome]
[Outcome]
A manager dashboard that sped up fraud resolution 2x, boosted team engagement, and gave managers clear oversight—turning data into actionable coaching strategies.
A manager dashboard that sped up fraud resolution 2x, boosted team engagement, and gave managers clear oversight—turning data into actionable coaching strategies.
[Secret Sauce]
[Secret Sauce]
Data-driven design + gamification that made work feel like a game—without sacrificing accuracy.
Data-driven design + gamification that made work feel like a game—without sacrificing accuracy.
[Process]
01. Problem Discovery
Conducted 8 user interviews to create personas, map user journeys, and identify key pain points in the workflow.
Conducted 8 user interviews to create personas, map user journeys, and identify key pain points in the workflow.
Conducted 8 user interviews to create personas, map user journeys, and identify key pain points in the workflow.
02. Design for Efficiency
Developed info. architecture, a conceptual model, prioritized features, ideated layouts, and tested mid-fi prototypes.
Developed info. architecture, a conceptual model, prioritized features, ideated layouts, and tested mid-fi prototypes.
Developed info. architecture, a conceptual model, prioritized features, ideated layouts, and tested mid-fi prototypes.
03. Design for Engagement
Integrated gamification mechanics linked to KPIs, improved visibility, and validated the final prototype same users.
Integrated gamification mechanics linked to KPIs, improved visibility, and validated the final prototype same users.
Integrated gamification mechanics linked to KPIs, improved visibility, and validated the final prototype same users.
[Act 1: Problem Discovery]
[Understanding current system workflow]
1.
1.
All Cases
All Cases
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Classified By
Classified By
2.
2.
AI Model
AI Model
Assigned to
Assigned to
3.
3.
Queues
Queues
Q1
Q1
Q2
Q2
Q3
Q3
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Resolved by
Resolved by
4.
4.
Analysts
Analysts
Manager's Responsibility
Manager's Responsibility
[The pain that was felt by the people using it every day]
Jessica
Fraud Ops Manager
Goal: Stop fraud faster while keeping her team motivated.
Pain: "I see 1,200 alerts but no patterns. How do I fix what I can’t see?"
Need: Jessica needed real-time actionable data and insights to make sure her team of analysts are performing.
Mark
Investigation Analyst
Goal: Make an impact, feel valued and not just process endless cases.
Pain: "100 cases a day, zero feedback—why does this feel pointless?"
Need: Mark needed to feel engaged, rewarded and grow in his role.
[Building a Dashboard for Jessica, the Manager]
I identified the essential objects Jessica needed to see —fraud queue progress, analyst productivity, and model performance—and actions she needed to take like assign, monitor and communicate.
Fraud Detection Manager
Fraud Detection Manager
Oversee
Oversee
Queue Progress
Queue Progress
Analyst’s Productivity
Analyst’s Productivity
Model Performance
Model Performance
Take Actions
Take Actions
Assign Analyst
Assign Analyst
Monitor Model Performance
Monitor Model Performance
Communicate With Team
Communicate With Team
[Act 2: Design for Efficiency]
[How Might We Help Jessica]
See the Big Picture
See the Big Picture
Why? To Track Multiple Objects Efficiently
Why? To Track Multiple Objects Efficiently
Overview Dashboard
Overview Dashboard
Queue
Queue
Analyst
Analyst
Analyst
Analyst
Model
Model
Alerts
Alerts
Why? Data Visualisation with Insights
Why? Data Visualisation with Insights
Make Informed Decisions
Make Informed Decisions
00
00
02
02
06
06
04
04
TRENDS
TRENDS
Threshhold
Threshhold
Prediction
Prediction
Cases / hr
Cases / hr
168
168
Mins / case
Mins / case
25m
25m
Why? Actionable Dashboard For Responding Quickly
Why? Actionable Dashboard For Responding Quickly
Take Essential Actions
Take Essential Actions
Export
Export
Assign
Assign
Overview Dashboard
Overview Dashboard
Queue
Queue
Alerts
Alerts
Model
Model
Analyst
Analyst
Action
Action
Assign
Assign
[Information Architecture]
I organized this data into overview, queues, analysts, and models, creating a logical flow that made insights easy to navigate.


[Ideation]
I explored multiple layouts—But only one made decisions effortless
Search
Search
Model
Model
Queue
Queue
Analyst
Analyst
Alerts
Alerts
Overview
Overview
Queue
Queue
5 Oct 2024
5 Oct 2024
Refresh
Refresh
Analyst
Analyst
Model
Model
Assign
Assign
Export
Export
PHNX
PHNX
Global
Objects
Global
Objects
Global Time
Controls
Global Time
Controls
Global CTA
Global CTA
Global Controls
Global Controls
Objects, Attributes
& analytics
Objects, Attributes
& analytics
Idea 1
Idea 1
Pros
Pros
Consistent Card Layout
No need to scroll back and forth
Consistent Card Layout
No need to scroll back and forth
Clear Reading Order
Alerts > Individual Cases
Clear Reading Order
Alerts > Individual Cases
Cons
Cons
Search
Search
Model
Model
Queue
Queue
Analyst
Analyst
Analyst
Analyst
Overview
Overview
Queue
Queue
5 Oct 2024
5 Oct 2024
Refresh
Refresh
Analyst
Analyst
Model
Model
Assign
Assign
Export
Export
PHNX
PHNX
Cases
Cases
Idea 2
Idea 2
Pros
Pros
Cons
Cons
Consistent Card Layout
No need to scroll back and forth
Consistent Card Layout
No need to scroll back and forth
Lack of emphasis and clear reading order
Cases is not priority to manager
Lack of emphasis and clear reading order
Cases is not priority to manager
Search
Search
Queue
Queue
Model
Model
Analyst
Analyst
Overview
Overview
Queue
Queue
5 Oct 2024
5 Oct 2024
Refresh
Refresh
Analyst
Analyst
Model
Model
Assign
Assign
Export
Export
PHNX
PHNX
Idea 3
Idea 3
Pros
Pros
Cons
Cons
Clear reading order
Consistent Card Layout
Clear reading order
Consistent Card Layout
Users might need to scroll back and forth
Side Nav Bar takes up real estate
Users might need to scroll back and forth
Side Nav Bar takes up real estate
[How I visualized data that drives action]
(Based on usability testing with 4 managers + 4 analysts with the prototype v01)
[Dsta Viz Purpose]
Comparison: Highlight fraud pattern shifts over time
Normalization: Filter noise to focus on key insights
Drill to Detail: Quick access to case specifics
Real-Time Alerts: Prompt managers to act on critical issues
Model FPR Tracking: Show AI failure rates for retraining
[Impact & Usability Test Results]
✅ "This would help me explain decisions to leadership." – Manager
✅ "Before, everything looked urgent. Now I know where to focus." – Analyst
✅ "I used to waste time digging for info—now it’s one click." – Manager
✅ Improved response time for high-risk fraud cases..
✅ Improved response time
for high-risk fraud cases.
✅ ""Helps me trust the system and adjust my workflow." – Analyst
✅ ""Helps me trust the system and adjust my workflow." – Analyst
[Prototype Version 01]
Swipe left to see the screens →
colour coded alerts for quick action
quick alerts
thresholds for quick scan
communicate faster with team
manager profile
Manager Overview
Take Actions
analyst overview
queue progress
overview
help in making informed decisions
Queue Detail
Queue Drill Down
Sorting and filtering data
speed focused and not quality driven
normalised data
colour coded
categorisation
performance
trends
Model Detail
Model Drill Down
model comparison & performance over time
comparison of case progress
drill down to specific model
case
attributes
colour coded alerts for quick action
quick alerts
thresholds for quick scan
communicate faster with team
manager profile
Manager Overview
Take Actions
analyst overview
queue progress
overview
help in making informed decisions
Queue Detail
Queue Drill Down
Sorting and filtering data
speed focused and not quality driven
normalised data
colour coded
categorisation
performance
trends
Model Detail
Model Drill Down
model comparison & performance over time
comparison of case progress
drill down to specific model
case
attributes
colour coded alerts for quick action
quick alerts
thresholds for quick scan
communicate faster with team
manager profile
Manager Overview
Take Actions
analyst overview
queue progress
overview
help in making informed decisions
Queue Detail
Queue Drill Down
Sorting and filtering data
speed focused and not quality driven
normalised data
colour coded
categorisation
performance
trends
Model Detail
Model Drill Down
model comparison & performance over time
comparison of case progress
drill down to specific model
case
attributes
colour coded alerts for quick action
quick alerts
thresholds for quick scan
communicate faster with team
manager profile
Manager Overview
Take Actions
analyst overview
queue progress
overview
help in making informed decisions
Queue Detail
Queue Drill Down
Sorting and filtering data
speed focused and not quality driven
normalised data
colour coded
categorisation
performance
trends
Model Detail
Model Drill Down
model comparison & performance over time
comparison of case progress
drill down to specific model
case
attributes
colour coded alerts for quick action
quick alerts
thresholds for quick scan
communicate faster with team
manager profile
Manager Overview
Take Actions
analyst overview
queue progress
overview
help in making informed decisions
Queue Detail
Queue Drill Down
Sorting and filtering data
speed focused and not quality driven
normalised data
colour coded
categorisation
performance
trends
Model Detail
Model Drill Down
model comparison & performance over time
comparison of case progress
drill down to specific model
case
attributes
colour coded alerts for quick action
quick alerts
thresholds for quick scan
communicate faster with team
manager profile
Manager Overview
Take Actions
analyst overview
queue progress
overview
help in making informed decisions
Queue Detail
Queue Drill Down
Sorting and filtering data
speed focused and not quality driven
normalised data
colour coded
categorisation
performance
trends
Model Detail
Model Drill Down
model comparison & performance over time
comparison of case progress
drill down to specific model
case
attributes
colour coded alerts for quick action
quick alerts
thresholds for quick scan
communicate faster with team
manager profile
Manager Overview
Take Actions
analyst overview
queue progress
overview
help in making informed decisions
Queue Detail
Queue Drill Down
Sorting and filtering data
speed focused and not quality driven
normalised data
colour coded
categorisation
performance
trends
Model Detail
Model Drill Down
model comparison & performance over time
comparison of case progress
drill down to specific model
case
attributes
colour coded alerts for quick action
quick alerts
thresholds for quick scan
communicate faster with team
manager profile
Manager Overview
Take Actions
analyst overview
queue progress
overview
help in making informed decisions
Queue Detail
Queue Drill Down
Sorting and filtering data
speed focused and not quality driven
normalised data
colour coded
categorisation
performance
trends
Model Detail
Model Drill Down
model comparison & performance over time
comparison of case progress
drill down to specific model
case
attributes
[While prototype 1 improved efficiency, it failed to engage users]
⚠️ "Day after day, it’s the same repetitive tasks—no feedback, no recognition. Why bother?" (90% of analysts)
⚠️ "We need more than just case counts—accuracy, retention, and productivity matter too." (80% of managers)
⚠️ "I can’t tell who’s excelling or who’s struggling without digging through reports." (Manager)
[Act 3: Design for Engagement]
[Turning work into a game — but one that drives productivity]
To drive engagement through competition, rewards, and progression.
Add real-time rewards (XP, badges) to break monotony and validate contributions.
Expand metrics to track quality, not just speed, giving managers deeper insights.
[The Flow]
Solve fraud
cases
Solve fraud
cases
Earn XP
& Coins
Earn XP
& Coins
Level
up
Level
up
Unlock
harder
cases
Unlock
harder
cases
Win Rewards &
Recognition
Win Rewards &
Recognition
Manager Journey
(Jessica)
Manager Journey
(Jessica)
Track
KPIs
Track
KPIs
Set game
objectives
Set game
objectives
Launch
Game
Launch
Game
Adjust
strategy
Adjust
strategy
Recognize & reward winners
Recognize & reward winners
[How I defined game mechanics, motivators, engagement loops, and incentives]
[Game Element]
XP & Levels
Leaderboards
Badges & Coins
Manager Tools
[Purpose]
Tracked progress via case resolution, and efficiency.
XP & Levels: Tracked progress via case resolution, and efficiency.
Ranked analysts by performance, fostering friendly competition.
Leadership: Ranked analysts by performance,
Rewarded accuracy and speed with virtual (and real) perks
Badges and Coins: Rewarded accuracy and speed with virtual (and real) perks
Let managers adjust rewards, track KPIs, and set targets.
Manager Tools: Let managers adjust rewards, track KPIs, and set targets.
[Impact & Usability Test Results]
✅ "Watching my level go up makes repetitive work feel meaningful." – Analyst
✅ "Leaderboard would transform how I manage my team's development."
✅ "I would be able to spend my team budget that actually drives growth and fraud reduction."
✅ "This would save me 2 hours a week in manual tracking and adjusting strategies." – Manager
[The final prototype, a system that finally worked for both teams]


Note: This is a GIF of the final prototype. Please give it a minute to load if you see a blank box. Thank you for your patience.
[Why it Worked]
✅ Analysts felt engaged and recognized—badges and XP made repetitive work meaningful, reducing burnout.
✅ Managers became strategic coaches—real-time KPIs helped cut weekly review time in half while improving team guidance.
✅ Healthy competition drove accuracy—leaderboards motivated analysts to resolve 20% more cases without sacrificing quality.
[Key Learnings]
Iterative testing pays off
Continuous user testing exposed blind spots, forcing a pivot from efficiency to engagement-driven design.
Iterative testing pays off
Continuous user testing exposed blind spots, forcing a pivot from efficiency to engagement-driven design.
Structure creates clarity
Clear IA and journey mapping transformed chaos into a streamlined, real-time workflow.
Structure creates clarity
Clear IA and journey mapping transformed chaos into a streamlined, real-time workflow.
Think in systems
Built interconnected tools that aligned analysts, managers, and business goals through shared data and incentives.
Think in systems
Built interconnected tools that aligned analysts, managers, and business goals through shared data and incentives.
Thank you for reading.
✨ Desktop recommended
This is the mobile-friendly summary—for deeper dive check out the complete case study on desktop!
Your interest means everything—thanks for exploring my work! 🌟"
[The pain that was felt by the people using it every day]
Jessica
Fraud Ops Manager
Goal: Stop fraud faster while keeping her team motivated.
Pain: "I see 1,200 alerts but no patterns. How do I fix what I can’t see?"
Need: Jessica needed real-time actionable data and insights to make sure her team of analysts are performing.
Mark
Investigation Analyst
Goal: Make an impact, feel valued and not just process endless cases.
Pain: "100 cases a day, zero feedback—why does this feel pointless?"
Need: Mark needed to feel engaged, rewarded and grow in his role.
[Building a Dashboard for Jessica, the Manager]
I identified the essential objects Jessica needed to see —fraud queue progress, analyst productivity, and model performance—and actions she needed to take like assign, monitor and communicate.
Fraud Detection Manager
Oversee
Queue Progress
Analyst’s Productivity
Model Performance
Take Actions
Assign Analyst
Monitor Model Performance
Communicate With Team
[How Might We Help Jessica]
See the Big Picture
Why? To Track Multiple Objects Efficiently
Overview Dashboard
Queue
Analyst
Analyst
Model
Alerts
Why? Data Visualisation with Insights
Make Informed Decisions
00
02
06
04
TRENDS
Threshhold
Prediction
Cases / hr
168
Mins / case
25m
Why? Actionable Dashboard For Responding Quickly
Take Essential Actions
Export
Assign
Overview Dashboard
Queue
Alerts
Model
Analyst
Action
Assign
[Information Architecture]
I organized this data into overview, queues, analysts, and models, creating a logical flow that made insights easy to navigate.

[How I visualized data that drives action]
(Based on usability testing with 4 managers + 4 analysts with the prototype v01)
[Data Viz Purpose]
Comparison - Highlight fraud pattern shifts over time.
Normalization - Filter noise to focus on key insights.
Drill to Detail - Quick access to case specifics
Real-Time Alerts - Prompt managers to act on critical issues
[Impact & Usability Test Results]
✅ 100% of managers gained better oversight
of trends
✅ Reduced cognitive load—managers
made faster decisions
✅ "I would be able to spend my team budget that actually drives growth and fraud reduction."
✅ "This would save me 2 hours a week in manual tracking and adjusting strategies." – Manager
[Defining the game mechanics, motivators, engagement loops, and incentives]
[Game Mechanics]
XP & Levels: Tracked progress via case resolution, and efficiency.
Leadership: Ranked analysts by performance,
Badges and Coins: Rewarded accuracy and speed with virtual (and real) perks
Manager Tools: Let managers adjust rewards, track KPIs, and set targets.
[Impact & Usability Test Results]
✅ "Watching my level go up makes repetitive work feel meaningful." – Analyst
✅ "Leaderboard would transform how I manage my team's development."
✅ "I would be able to spend my team budget that actually drives growth and fraud reduction."
✅ "This would save me 2 hours a week in manual tracking and adjusting strategies." – Manager
[Why it Worked]
✅ Analysts felt engaged and recognized—badges and XP made repetitive work meaningful, reducing burnout.
✅ Managers became strategic coaches—real-time KPIs helped cut weekly review time in half while improving team guidance.
✅ Healthy competition drove accuracy—leaderboards motivated analysts to resolve 20% more cases without sacrificing quality.
[Key Learnings]
Iterative testing pays off
Continuous user testing exposed blind spots, forcing a pivot from efficiency to engagement-driven design.
Structure creates clarity
Clear IA and journey mapping transformed chaos into a streamlined, real-time workflow.
Think in systems
Built interconnected tools that aligned analysts, managers, and business goals through shared data and incentives.
[The final prototype, a system that finally worked for both teams]

[Outcome]
🖥️ Desktop recommended
This is the condensed mobile version - swing by my desktop site for the full case study details! Thanks for checking it out—you're awesome! ✨