B2B

Saas

Web app

Digitizing a process that couldn't be simplified, only translated. 

Designed a fraud investigation platform used by a 30-person team, transforming an unstructured process into a centralized system with clear ownership, prioritization, and traceability.

Led UX strategy and end-to-end product design, structuring the complex investigation workflow into a single system with full visibility for users.

Result: 4,000+ cases centralized, faster investigation cycles, and zero information loss across the workflow.

Role

Product Designer. End-to-end.

timeline

12 months

Team

PO + Tech Lead + 15 devs

impact

4000+ centralized cases, clarified actions, optimized investigations

Context

Quálitas’ fraud team handled over 4,000 ongoing investigations through a fully manual process involving emails, calls, messaging apps, and spreadsheets.

This fragmentation made cases difficult to track, slowed down decision-making, and led to lost information and financial risk.

How might we translate this manual process into a digital one?

Some tools formerly used during investigations

Problem framing

Analysts lacked a reliable way to understand what required action at any given moment.

The workflow was long and complex, and with a small team handling thousands of cases, analysts had no reliable way to know which case required action at any given time.

Analysts need clarity and visibility to know when to act, so investigations can be completed faster.

KEY DECISION

My role

I owned the end-to-end design process, from discovery sessions with analysts to final handoff.

I worked alongside the Product Owner and the Tech Lead on product decisions. I led the discovery sessions, learned and understood the workflow, and presented design decisions with the team, iterating with direct feedback. I also led handoff for the dev team, explaining the whole system and how each workflow section worked to ensure the team understood and translated each function as it had to work.

Constraints & tradeoffs

Working within boundaries

The workflow was predefined due to operational and compliance constraints, meaning we couldn’t restructure the process.

I focused on making the workflow understandable, traceable, and actionable for each role, reducing friction without altering the existing workflow.

The workflow was predefined due to operational and compliance constraints, meaning we couldn’t restructure the process.

I focused on making the workflow understandable, traceable, and actionable for each role, reducing friction without altering the existing workflow.

Given a 12-month deadline, I prioritized delivery speed over system flexibility.

To move faster, I collaborated with the Tech Lead to design key actions in modals instead of full pages. This reduced the number of screens and sped up delivery, while accepting limited flexibility in some flows.

Given a 12-month deadline, I prioritized delivery speed over system flexibility.

To move faster, I collaborated with the Tech Lead to design key actions in modals instead of full pages. This reduced the number of screens and sped up delivery, while accepting limited flexibility in some flows.

SYSTEM OVERVIEW

The platform supports the full investigation lifecycle across three roles.

I prioritized that each claim detail had the most important information clear and visible from the get-go, all information categorized and divided in different tables to make it easier to read, each case status and timeline clear, and actionable CTAs prioritized in the structure.

Design decisions

Key choices

Starting with high fidelity

a

a

a

a. Action Centre

The platform couldn't expect analysts to track 4,000+ cases manually. It had to tell them what needed attention.

With thousands of active cases on the platform, a standard case list wasn’t enough to track critical cases. Analysts couldn't be expected to remember pending tasks across hundreds of open investigations.

Introducing prioritization meant analysts had to rely on the system instead of their own tracking methods, so clarity and trust became critical.

a. Action Centre

The platform couldn't expect analysts to track 4,000+ cases manually. It had to tell them what needed attention.

With thousands of active cases on the platform, a standard case list wasn’t enough to track critical cases. Analysts couldn't be expected to remember pending tasks across hundreds of open investigations.

Introducing prioritization meant analysts had to rely on the system instead of their own tracking methods, so clarity and trust became critical.

b. Skipping Lo-Fi

Skipping lo-fi was the fastest way to validate a system this complex.

With a 12-month deadline and a complex workflow, I used interactive prototyping to quickly explore and align functionalities within the platform. This approach prioritized speed over depth, but it allowed us to validate quickly.

A functioning prototype helped analysts interact with the design and test ideas early, reducing ambiguity before development.

b. Skipping Lo-Fi

Skipping lo-fi was the fastest way to validate a system this complex.

With a 12-month deadline and a complex workflow, I used interactive prototyping to quickly explore and align functionalities within the platform. This approach prioritized speed over depth, but it allowed us to validate quickly.

A functioning prototype helped analysts interact with the design and test ideas early, reducing ambiguity before development.

c. invoice system

Catching errors before they reach a reviewer requires system-wide validation logic.

Before, reimbursements for investigators were tracked through email chains and it was easy to lose the thread and delay payments. I found rejections were common, mostly from expenses exceeding approved limits.

I defined automatic validations so the system catches errors before they reach a reviewer. If an expense goes over the threshold, it's rejected immediately. Strict validations reduced errors, but also meant rejecting inputs earlier, which could feel restrictive at first.

c. invoice system

Catching errors before they reach a reviewer requires system-wide validation logic.

Before, reimbursements for investigators were tracked through email chains and it was easy to lose the thread and delay payments. I found rejections were common, mostly from expenses exceeding approved limits.

I defined automatic validations so the system catches errors before they reach a reviewer. If an expense goes over the threshold, it's rejected immediately. Strict validations reduced errors, but also meant rejecting inputs earlier, which could feel restrictive at first.

"Montse understands requirements quickly and turns complex flows into clear, functional solutions. Her strategic mindset combined with empathy and collaboration skills makes everything flow effortlessly"

Alejandra Zacarías

Product Owner at The Rocket Code

"Montse understands requirements quickly and turns complex flows into clear, functional solutions. Her strategic mindset combined with empathy and collaboration skills makes everything flow effortlessly"

Alejandra Zacarías

Product Owner at The Rocket Code

Operational impact

Streamlining Fraud Investigations

4,000+ cases centralized into a single system
Eliminated reliance on external tools (email, spreadsheets, messaging)
Improved investigation speed through prioritization and validation
Zero information loss across the investigation lifecycle

Learnings

Learning a domain fast is a design skill.

I had no background in insurance or fraud, so I worked closely with analysts to understand their process and translate it into a system they could use.

Knowing where to invest design effort

With a complex system with an extensive workflow, I had to decide what I could optimize better and what I had to make work for the first release. Understanding what to preserve was as important as what to change.

Alignment is as important as the solution

Collaboration with Engineering wasn’t always smooth. Relying only on handoffs created gaps in understanding. I learned that working directly with developers and explaining the full workflow and edge cases improved our collaboration greatly.

reflection

I’d explore how analysts actually use the Action Center over time, where cases stall, and how prioritization impacts resolution. There are also parts of the workflow that could be improved, especially where the process tightens. Due to time constraints, I focused on translating the workflow faithfully, but I’d revisit these areas to explore more flexible solutions.

Interested in collaborating?

Say hello.

click to copy!

Open to Product Design roles, remote, full-time.

Montse Mundo © 2026

Interested in collaborating?

Say hello.

Tap to copy

Open to Product Design roles, remote, full-time.

Montse Mundo © 2026

Interested in collaborating?

Say hello.

Tap to copy

Open to mid/senior Product Design roles: remote, full-time, EU & Americas timezones

Montse Mundo © 2026