Product Designer for enterprise software. Product Designer
for enterprise software.

In critical flows, I make the intent and consequence of each step visible before the user acts. On SAJ Online, I applied this to extrajudicial hearings and to converting procedures into court cases. For a court-integration status panel, I drafted ten design conclusions with the UX researcher. Two of them: show status per action, and warn before an action fails. I presented the draft to engineering, and the panel was not built.

Feature design · Ordering and support automation · Company website

The Maria da Penha Dashboard started from the previous system’s structure, so prosecutors kept familiar patterns in a sensitive domain. On Bússola, at Grupo Boticário, executives worked in Excel, so I used spreadsheet patterns: tabs, filters and cells. When users read correct numbers in different ways, I added a time period to every metric and help text where they met it.

Decision dashboard · Feature design · Online store

My MBA thesis in Digital Business (USP/Esalq, graded 10/10) asked why people pay for digital content. Support was more consistent when the content was useful in practice and had a community around it. The barriers were expecting free access, low perceived value and distrust of the platform. In client work, I rank findings by the metric you track: sales, contacts or retention. At Elo, I compared scroll and click behavior on the Use Elo mobile pages, then reworked banners, CTAs and hierarchy. The changes contributed to a 15% increase in user retention and a higher registration rate.

Website and conversion audit · Landing page · Product strategy

Prosecutors testing Softplan’s AI workspace already had a licensed general-purpose AI assistant. One participant compared both tools himself: the legal content was similar, and speed decided his preference. Across sessions, users asked for legal knowledge, sources they could check and a tool that works every time. I turned this into requirements: a visible generation process, linked references and case law users can verify.

Product strategy · Feature design · Product UX diagnostic

For the AI workspace, I proposed one guiding metric and a few supporting indicators. I wrote the tracking plan for the core loop: open, start generation, complete, export. In soft launch, it showed that 1 of 12 case imports finished at one client office. On SAJ Online, success means users complete critical actions without errors. I read open-text CSAT comments by the usability problem they describe, because the score alone hides what users meant.

Telemetry plan · Monthly evolution · Launch

Each session has a goal, a participant profile and tasks built around the actions that deliver value. I review recordings, group evidence by how often it repeats and rate each finding on Nielsen’s severity scale. In AI products, I test navigation and output quality as separate tasks. One participant navigated without difficulty and rejected the generated content.

User research · Feature design · Company website (design tested with five customers)

I sent per-client usage summaries of the AI workspace to the head of the business unit. He replied by defining four minimum indicators for the team. When a design concern is overruled, I record what was proposed, what was cut and the expected consequence. At Franklin Templeton, I handed over interaction specs, component libraries and design tokens to distributed engineering teams.

Product strategy · Design system audit · Monthly evolution