From Bureaucracy to Trust

The starting point

In 2015 I was an HR analyst at Caixa Seguradora, and one of my responsibilities was the employee benefits reimbursement process. It was entirely manual: branches spread across the country, paper receipts, endless email threads and a different point of contact for each benefit category. I watched what that caused from both sides of the counter.

On the employee side, nobody knew whether a request had been received, who was handling it or when the money would land. Communication delays became payment delays, and payment delays hit the personal finances of people who were simply trying to use a benefit they were entitled to. When an exception was granted for one case and denied for a similar one, a perception of unfairness took root, the kind of damage no apology email can undo.

On the HR side, the People team was drowning in repetitive work: checking, chasing, resending, filling spreadsheets. There was no searchable history, no duplicate prevention, no clean data for strategic analysis. Compliance pressure kept growing without any tooling to support it: duplicate payments, fraud exposure and internal audit risk were real scenarios, not hypotheticals. Add the hours of operational work that generated no value at all and the hidden cost of the process became impossible to ignore.

The challenge

Turn that manual flow into a digital process with speed and transparency from end to end, serving three audiences at once: employees, who needed predictability; HR, who needed to move from operational grind to analytical work; and the company, which needed governance and traceability.

Why SharePoint

Because it was the tool we already had in the building. There was no budget for a new system and no runway for a long implementation, but there was SharePoint, approved by IT, integrated with the corporate login and capable of supporting validated forms, approval flows and consolidated dashboards. Choosing the available tool and designing well on top of it was the project's first decision, and one I still stand behind: technical constraints don't prevent a good experience, they just demand sharper design judgment.

The journey

I wasn't calling any of this UX yet. At the time I applied, through intuition and analytical reasoning, what years later I would recognize as UX Strategy: before proposing a single screen, I went to understand how the process actually worked, not how the manual said it worked.

I mapped the full reimbursement journey twice, once through the employee's eyes and once through HR's. On paper, the two journeys told the same story from different angles: the employee gathered receipts, figured out who to email depending on the benefit, sent messages and waited in the dark; HR received requests through scattered channels, checked everything by hand, answered the same questions over and over and could never see the whole picture.

That mapping, fed by interviews and collaborative workshops with HR, IT and Internal Communications, pinned each pain point to its exact stage. It turned loose complaints into a diagnosis the three areas could read together and, more importantly, prioritize together.

For the Flow Benefits case, I recreated that era's system as a navigable simulation. You can walk the same path the employee faced, from the category menu to the error that interrupted the request halfway through.

The three screens below mark the moments the mapping turned into a diagnosis.

Process and role

My role was to represent the voice of the People team and the business, coordinating HR, IT and Internal Communications in an agile, collaborative effort. In practice, the method had five phases:

1. Listening and research. I ran interviews and workshops with the three areas to understand flows, exceptions and friction points, from request to payment.
2. Journey mapping. I drew the employee and HR journeys step by step, identifying the pain at each moment and the cost of each round of rework.
3. Process redesign. I modeled the end-to-end flow with standardized rules and automated validity and eligibility checks, taking away from people the decisions a rule could make on its own.
4. Forms and microcopy. I simplified the forms with validation logic to cut inconsistencies at the source, and wrote every instruction and message in the flow to eliminate doubt before it turned into an email. Two of Nielsen's heuristics guided everything: error prevention and visibility of system status.
5. Rollout with IT. I brought the design into SharePoint alongside the technical team, with a responsive interface that let employees open a request from their phone and photograph the receipt on the spot, no paper, no scanner.

The delivery

The before and after each fit in one sentence. Before: data scattered across emails and files, constant rework and error risk at every manual step. After: a centralized digital process with document upload, data consolidated into dashboards and status feedback to the employee along the entire way.

The results showed up on all three fronts the challenge demanded. On operational efficiency, rework dropped and the hours once spent on repetitive tasks went back to what mattered: people analytics and strategy. On governance, the process gained full traceability, with lower fraud risk and the conditions to actually pass an audit. And on employee experience, reimbursement became transparent and predictable: satisfaction with the process grew, and with it, trust in the employer brand.

Learnings

1. Internal processes are user experiences. That sentence became a career principle. Information architecture and UX writing didn't just make the system functional: they built trust between people and company, which was what was actually broken.
2. I was doing UX before I knew its name. Feedback from an IT colleague about my focus on experience design was the first external validation that this way of working had a name, a method and a market. I look at this project as the day I discovered my profession.
3. Trust is built on visibility and consistency. Employees weren't asking for speed above all else; they wanted to know where their request stood and to see the same rule applied to everyone. Clear status and standardized rules solved more than any deadline promise could.
4. Good problems keep paying off. Ten years later I revisited this exact problem with my current toolkit: I prototyped the solution in Figma with AI-powered automations for microcopy and validation, and it evolved into Flow Benefits, the case that connects my Employee Experience roots to end-to-end product delivery.