UX Research
Antifraud System
Anti-Fraud System redesign focused on UX research and strategy; addressing slow detection and security gaps by mapping workflows, benchmarking, and defining opportunities for automation and real-time monitoring.
Year :
2025
Industry :
Public Transportation
Client :
Confident
Project Duration :
3 months
Context:
I led the UX research and discovery for a large-scale public transportation card system with the goal of designing a preventive, scalable anti-fraud platform. The system handled multiple card types (individual, corporate, gratuity, and tourist cards) and fraud was being detected reactively, through manual processes. The business objective was clear: reduce fraud significantly in the early stages of the customer journey while ensuring the solution could scale and eventually evolve into a potential SaaS product.
Problem:
Fraud detection relied heavily on manual spreadsheets, ad-hoc scripts, and the tacit knowledge of a very small team. Cases took 5–10 business days to be resolved, false positives often penalized legitimate customers, and frontline teams (such as stores and call centers) lacked visibility or autonomy to provide answers. In short, the system had data but no centralized, automated or user-friendly way to transform that data into fast, reliable decisions.
Approach:
I structured the engagement into two phases: a rapid discovery to uncover root causes and a strategic framing to design the solution.
Discovery & Evidence Gathering – Desk research, service flow analysis, and a CSD (Certainties, Assumptions, Doubts) workshop grounded the interviews and reduced bias. This revealed three operational pillars used to detect fraud: registration, recharge, and utilization.
User Research – Semi-structured interviews with analysts, store attendants, call center agents and developers surfaced key pain points: lack of automation, fragmented information, no test environment for rules, and poor communication across teams.
Personas & Journeys – I created personas representing different roles (Analyst, Store Attendant, Call Center Agent, Developer) and mapped their journeys. This exposed inefficiencies such as repetitive tasks, false positives, and operational bottlenecks.
Benchmarking – I researched market leaders in fraud prevention (e.g. SaaS solutions and APIs for behavioral analysis, biometrics, rules engines) to understand feasible integrations and cost/benefit trade-offs.
Definition & Roadmap – I translated insights into opportunity areas: automation for triage, a unified operations dashboard, role-based access, a sandbox for testing detection rules, and audit logs. These were prioritized in a roadmap structured around PoC → MVP → scale.
Personas:
Fraud Analyst: Needs automation and a consolidated view to reduce manual workload.
Store Attendant: Needs clear explanations and guided steps to assist customers quickly.
Call Center Agent: Needs real-time status visibility to resolve cases without escalation.
Developer: Needs standardized requests and a sandbox to implement rules safely.
Solution:
The proposed design included:
A centralized dashboard with prioritized alerts, historical views, and actionable decisions (block, escalate, whitelist).
Automated triage rules to cluster suspicious behaviors while keeping a human-in-the-loop for validation.
Simplified, role-based UIs for frontline staff, providing plain-language reasons and guided scripts.
A rules sandbox with version control to test new detection logic without jeopardizing operations.
Integrated logging and audit trails for accountability and continuous learning.
Each design element was informed by UX strategy, information architecture, interaction design, data visualization, and usability testing, ensuring the experience was both powerful for analysts and accessible for frontline teams.
Impact:
The redesign created measurable business impact:
Reduction of more than R$100,000 in monthly financial losses through faster fraud detection and prevention.
Decrease in resolution time from 5–10 business days to near real time.
Fewer false positives, improving trust and reducing friction for legitimate customers.
Greater autonomy for frontline teams, minimizing repeated store visits and customer frustration.
Foundation for a scalable product roadmap, with potential to evolve into a SaaS anti-fraud solution.
Key Takeaways:
This project highlights my ability as a UX/UI Designer to conduct user-centered discovery in complex operational contexts, translate qualitative insights into product strategy, and design pragmatic, actionable solutions. By combining UX research, journey mapping, benchmarking, information architecture, and high-fidelity prototyping, I created a roadmap that balanced business impact, technical feasibility, and user experience.
More Projects
hey!
hey!
hey!
Let’s talk design
Let’s talk design
Let’s talk design
I’m open to new projects, collaborations, or job opportunities. send me a message and let’s chat!
I’m open to new projects, collaborations, or job opportunities. send me a message and let’s chat!
I’m open to new projects, collaborations, or job opportunities. send me a message and let’s chat!
UX Research
Antifraud System
Anti-Fraud System redesign focused on UX research and strategy; addressing slow detection and security gaps by mapping workflows, benchmarking, and defining opportunities for automation and real-time monitoring.
Year :
2025
Industry :
Public Transportation
Client :
Confident
Project Duration :
3 months
Context:
I led the UX research and discovery for a large-scale public transportation card system with the goal of designing a preventive, scalable anti-fraud platform. The system handled multiple card types (individual, corporate, gratuity, and tourist cards) and fraud was being detected reactively, through manual processes. The business objective was clear: reduce fraud significantly in the early stages of the customer journey while ensuring the solution could scale and eventually evolve into a potential SaaS product.
Problem:
Fraud detection relied heavily on manual spreadsheets, ad-hoc scripts, and the tacit knowledge of a very small team. Cases took 5–10 business days to be resolved, false positives often penalized legitimate customers, and frontline teams (such as stores and call centers) lacked visibility or autonomy to provide answers. In short, the system had data but no centralized, automated or user-friendly way to transform that data into fast, reliable decisions.
Approach:
I structured the engagement into two phases: a rapid discovery to uncover root causes and a strategic framing to design the solution.
Discovery & Evidence Gathering – Desk research, service flow analysis, and a CSD (Certainties, Assumptions, Doubts) workshop grounded the interviews and reduced bias. This revealed three operational pillars used to detect fraud: registration, recharge, and utilization.
User Research – Semi-structured interviews with analysts, store attendants, call center agents and developers surfaced key pain points: lack of automation, fragmented information, no test environment for rules, and poor communication across teams.
Personas & Journeys – I created personas representing different roles (Analyst, Store Attendant, Call Center Agent, Developer) and mapped their journeys. This exposed inefficiencies such as repetitive tasks, false positives, and operational bottlenecks.
Benchmarking – I researched market leaders in fraud prevention (e.g. SaaS solutions and APIs for behavioral analysis, biometrics, rules engines) to understand feasible integrations and cost/benefit trade-offs.
Definition & Roadmap – I translated insights into opportunity areas: automation for triage, a unified operations dashboard, role-based access, a sandbox for testing detection rules, and audit logs. These were prioritized in a roadmap structured around PoC → MVP → scale.
Personas:
Fraud Analyst: Needs automation and a consolidated view to reduce manual workload.
Store Attendant: Needs clear explanations and guided steps to assist customers quickly.
Call Center Agent: Needs real-time status visibility to resolve cases without escalation.
Developer: Needs standardized requests and a sandbox to implement rules safely.
Solution:
The proposed design included:
A centralized dashboard with prioritized alerts, historical views, and actionable decisions (block, escalate, whitelist).
Automated triage rules to cluster suspicious behaviors while keeping a human-in-the-loop for validation.
Simplified, role-based UIs for frontline staff, providing plain-language reasons and guided scripts.
A rules sandbox with version control to test new detection logic without jeopardizing operations.
Integrated logging and audit trails for accountability and continuous learning.
Each design element was informed by UX strategy, information architecture, interaction design, data visualization, and usability testing, ensuring the experience was both powerful for analysts and accessible for frontline teams.
Impact:
The redesign created measurable business impact:
Reduction of more than R$100,000 in monthly financial losses through faster fraud detection and prevention.
Decrease in resolution time from 5–10 business days to near real time.
Fewer false positives, improving trust and reducing friction for legitimate customers.
Greater autonomy for frontline teams, minimizing repeated store visits and customer frustration.
Foundation for a scalable product roadmap, with potential to evolve into a SaaS anti-fraud solution.
Key Takeaways:
This project highlights my ability as a UX/UI Designer to conduct user-centered discovery in complex operational contexts, translate qualitative insights into product strategy, and design pragmatic, actionable solutions. By combining UX research, journey mapping, benchmarking, information architecture, and high-fidelity prototyping, I created a roadmap that balanced business impact, technical feasibility, and user experience.
More Projects
hey!
hey!
hey!
Let’s talk design
Let’s talk design
Let’s talk design
I’m open to new projects, collaborations, or job opportunities. send me a message and let’s chat!
I’m open to new projects, collaborations, or job opportunities. send me a message and let’s chat!
I’m open to new projects, collaborations, or job opportunities. send me a message and let’s chat!
UX Research
Antifraud System
Anti-Fraud System redesign focused on UX research and strategy; addressing slow detection and security gaps by mapping workflows, benchmarking, and defining opportunities for automation and real-time monitoring.
Year :
2025
Industry :
Public Transportation
Client :
Confident
Project Duration :
3 months
Context:
I led the UX research and discovery for a large-scale public transportation card system with the goal of designing a preventive, scalable anti-fraud platform. The system handled multiple card types (individual, corporate, gratuity, and tourist cards) and fraud was being detected reactively, through manual processes. The business objective was clear: reduce fraud significantly in the early stages of the customer journey while ensuring the solution could scale and eventually evolve into a potential SaaS product.
Problem:
Fraud detection relied heavily on manual spreadsheets, ad-hoc scripts, and the tacit knowledge of a very small team. Cases took 5–10 business days to be resolved, false positives often penalized legitimate customers, and frontline teams (such as stores and call centers) lacked visibility or autonomy to provide answers. In short, the system had data but no centralized, automated or user-friendly way to transform that data into fast, reliable decisions.
Approach:
I structured the engagement into two phases: a rapid discovery to uncover root causes and a strategic framing to design the solution.
Discovery & Evidence Gathering – Desk research, service flow analysis, and a CSD (Certainties, Assumptions, Doubts) workshop grounded the interviews and reduced bias. This revealed three operational pillars used to detect fraud: registration, recharge, and utilization.
User Research – Semi-structured interviews with analysts, store attendants, call center agents and developers surfaced key pain points: lack of automation, fragmented information, no test environment for rules, and poor communication across teams.
Personas & Journeys – I created personas representing different roles (Analyst, Store Attendant, Call Center Agent, Developer) and mapped their journeys. This exposed inefficiencies such as repetitive tasks, false positives, and operational bottlenecks.
Benchmarking – I researched market leaders in fraud prevention (e.g. SaaS solutions and APIs for behavioral analysis, biometrics, rules engines) to understand feasible integrations and cost/benefit trade-offs.
Definition & Roadmap – I translated insights into opportunity areas: automation for triage, a unified operations dashboard, role-based access, a sandbox for testing detection rules, and audit logs. These were prioritized in a roadmap structured around PoC → MVP → scale.
Personas:
Fraud Analyst: Needs automation and a consolidated view to reduce manual workload.
Store Attendant: Needs clear explanations and guided steps to assist customers quickly.
Call Center Agent: Needs real-time status visibility to resolve cases without escalation.
Developer: Needs standardized requests and a sandbox to implement rules safely.
Solution:
The proposed design included:
A centralized dashboard with prioritized alerts, historical views, and actionable decisions (block, escalate, whitelist).
Automated triage rules to cluster suspicious behaviors while keeping a human-in-the-loop for validation.
Simplified, role-based UIs for frontline staff, providing plain-language reasons and guided scripts.
A rules sandbox with version control to test new detection logic without jeopardizing operations.
Integrated logging and audit trails for accountability and continuous learning.
Each design element was informed by UX strategy, information architecture, interaction design, data visualization, and usability testing, ensuring the experience was both powerful for analysts and accessible for frontline teams.
Impact:
The redesign created measurable business impact:
Reduction of more than R$100,000 in monthly financial losses through faster fraud detection and prevention.
Decrease in resolution time from 5–10 business days to near real time.
Fewer false positives, improving trust and reducing friction for legitimate customers.
Greater autonomy for frontline teams, minimizing repeated store visits and customer frustration.
Foundation for a scalable product roadmap, with potential to evolve into a SaaS anti-fraud solution.
Key Takeaways:
This project highlights my ability as a UX/UI Designer to conduct user-centered discovery in complex operational contexts, translate qualitative insights into product strategy, and design pragmatic, actionable solutions. By combining UX research, journey mapping, benchmarking, information architecture, and high-fidelity prototyping, I created a roadmap that balanced business impact, technical feasibility, and user experience.
More Projects
hey!
hey!
hey!
Let’s talk design
Let’s talk design
Let’s talk design
I’m open to new projects, collaborations, or job opportunities. send me a message and let’s chat!
I’m open to new projects, collaborations, or job opportunities. send me a message and let’s chat!
I’m open to new projects, collaborations, or job opportunities. send me a message and let’s chat!