Reducing friction in FGC deposit insurance fund recovery
Time
5 months - April to August 2026
Domain Area
Mobile Aplication
My role
Product Designer
Company
FGC - Brazil's Deposit Guarantee Fund
Tools
Figma, Figjam, Figma Make, Maze, Gemini, Antigravity, ChatGPT Images 2.0
Disclaimer
Project initiated as my Capstone Project (TCC) at PUC-RS and developed independently based on real research data, without formal affiliation with FGC.
Executive Summary
The FGC mobile app is the sole digital channel in Brazil for customers to claim deposit insurance reimbursements of up to R$ 250,000 following Central Bank financial institution liquidations. The legacy experience suffered from system downtime, technical jargon, identity verification friction, and a lack of status transparency during high-stress moments. I conducted research and redesigned core user journeys, focusing on progressive disclosure, transparent status tracking, and simplified claim submission.
Achieved impact:
100% task success rate
In unmoderated usability testing (19 participants)
Positive qualitative feedback
Praising transparency in claim status tracking
Expected impact:
Reduced effort in fund recovery
Guided, decluttered flows to minimize cognitive load.
Decrease in support ticket volume
Clear visibility into guarantee status and processing timelines to reduce user inquiries.
Autonomy and emotional security
Actionable copy during system errors to restore user control and confidence.
Project context
The FGC app is the only digital channel in Brazil where customers can request reimbursement of up to R$ 250,000 following the liquidation of financial institutions by the Central Bank. The existing experience suffered from instability, confusing language, friction during identity verification, and a lack of transparency regarding status and timelines—causing frustration during an already stressful time. I conducted research and redesigned the user journeys, focusing primarily on providing clear guidance, status transparency, and simplifying the request process.
How might we transform a high-stress asset recovery journey into a transparent, reassuring, and intuitive digital experience?
Financial liquidation triggers immediate anxiety. As the single digital channel for asset recovery, the app must guide users with complete predictability—ensuring the claiming process provides relief rather than additional friction.
Learning about users and the journey
To uncover root causes and understand user behavior in this critical journey, I structured a multi-method Discovery combining qualitative, quantitative research, and an AI-assisted heuristic audit.
Desk Research
Researched deposit guarantee fund frameworks and benchmarked international recovery platforms.
Analysis of app store reviews
Mapped key user friction points by analyzing 516 authentic app store reviews
AI-Agent Assisted Heuristic Evaluation
Audited 18 core screens using a custom Human-In-the-Loop Multi-Agent AI system I built.
Mental Model Mapping
Mapped user expectations, fears, and mental models during stressful fund recovery moments.
Key insights
Reframing insights into opportunities & prioritization
With the research data in hand, I revisited all the identified opportunities to brainstorm solutions and filter for initiatives with the highest user impact and lowest implementation complexity.
How Might We...
I applied this technique to transform the pain points and opportunities identified in the research into solution-generating questions.
Effort vs Impact Matrix
I used the matrix to prioritize the proposals, focusing my efforts on opportunities in the "Do Now" quadrant, which consists of high-impact, low-effort tasks.
Structuring design hypotheses
To guide design decisions and the construction of the flows, I structured four main hypotheses:
Exploring proposals from paper to high fidelity
To validate structural paths with agility, I combined manual sketches and AI-assisted exploration to iterate quickly before building the final high-fidelity version.
Scribbling concepts on paper
I mapped the initial structure of flows and screens in low fidelity to test layout alternatives without visual attachment.
Rapid prototyping with AI (Figma Make / Antigravity)
I generated layouts and screen variations quickly to illustrate and test different navigation paths.
Generation of visual assets (ChatGPT Images 2.0)
I created supporting illustrations and visual elements for the project, ensuring aesthetic consistency.
UX Writing refinement with AI
I used Skills to review textual clarity, simplify technical jargon, and build an accessible vocabulary.
First high-fidelity version
I built prototypes of the main journeys to test hypotheses directly with users.
Validating usability solutions with users
I conducted a round of unmoderate remote tests on the Maze platform with 19 participants. Participants performed 3 main tasks:
Find a way to get their money back.
Confirm identity.
Track request status.
Results obtained:
100% of task success
In usability testing (19 participants)
Positive qualitative feedback
Regarding clarity in tracking the request
Key findings:
First-click dispersion
8 out of 19 users missed the first click on the refund request task.
Similar naming
2 out of 19 users noted that some features had very similar names.
Acknowledgment of the term "Warranty"
Recognition of the term "Guarantee" 2 out of 19 users mentioned that the term "guarantee" on its own was not immediately recognized as the way to get their money back.
Refining solutions based on lessons learned
Problem: critical access errors and technical language
System errors displayed technical jargon without offering solutions, blocking app access and causing panic during a moment of financial stress.
Antes
Solution: creating humanized, helpful error messages
I redesigned the error messages using plain language and practical workarounds to restore the user's peace of mind and sense of control.
After
Problem: lack of direction before guarantee release
The initial screen offered no useful actions while the guarantee was pending release, leaving the user without guidance on what to do during the wait.
Antes
Solution: pre-registration of the receiving account
I introduced an option to pre-register a bank account, ensuring it was ready for when the guarantee request was approved.
Novo
Problem: lack of transparency in request tracking
The absence of expected timeframes and superficial tracking created anxiety and uncertainty regarding when the funds would be deposited.
Before
Solution: detailed tracking with request timelines
I optimized request tracking by detailing the stages and displaying estimated timeframes for fund recovery, aiming to convey transparency and provide emotional reassurance.
Depois
Problem: friction and rejections during document validation
A lack of step-by-step guidance and the inability to replace uploaded files increased user effort and the risk of submitting illegible photos.
Solution: simplified instructions and quality checks
I structured guided instructions and added a document legibility check prior to submission to ensure photo quality and increase approval rates.
New
Impact achieved
100% success rate
In usability testing (19 participants)
Positive qualitative feedback
Regarding clarity in request tracking
Expected impact:
Reduced effort during redemption
Guided, simplified flows to lower the user's cognitive load.
Reduction in support calls
Clear visibility into warranty status and timelines to minimize uncertainty.
Autonomy and peace of mind
Clear guidance during errors and instability to restore a sense of control to the user.
