Rethinking a Workflow from the Ground Up.

How a mid-stream pivot and a two-click demo reframed the way a team thought about the future of their work.

01. About the Project

This one isn't a typical case study. There are no polished wireframe galleries or A/B test results. What this project does have is a story about catching a major inflection point mid-stream, making a hard call to pivot, and using design thinking to reframe how an entire team understood the future of their work.


My research / service design partner and I were brought in to lead the workflow modernization for a financial stress management process. It’s a critical, highly regulated function that had long lived in spreadsheets and offline tools. The goal was to bring it online: build it into the platform, keep it auditable, and make it sustainable. Standard enough.


Then agentic AI entered the conversation, and nothing about our approach stayed standard.

02. The challenge

We were designing for a small team of specialists doing deeply judgment-heavy work in a high-compliance environment. Their process was intricate, they spent their days manually identifying anomalies in system-generated data, investigating the causes, weighing options, making adjustments, and documenting every decision. Institutional knowledge was everywhere. Very little of it was captured anywhere.


The technology initiative was focused on consolidating this process into a single platform — a real step forward from the fragmented, spreadsheet-dependent status quo. Our job was to understand the current-state workflow deeply enough to design something that actually served the people doing the work.
Discovery was its own challenge. This team had a long history of being handed technology solutions that didn't reflect their feedback. Trust was low, and anxiety about automation was real. A small, specialized team in a high-efficiency push has good reason to wonder what "modernization" means for their jobs.

03. The Pivot

We had identified the happy path, a clean initial use case to anchor the first design around, when we had to stop and ask ourselves a hard question.


Everything we were designing was a digital version of what users already did manually. Search for anomalies. Investigate. Decide. Adjust. Sign off. The same cognitive load in a cleaner interface was an improvement, but hardly a transformation.


Agentic AI flipped all of that on its head. Once we saw what was possible, it became clear that continuing down the traditional design path meant building something we'd have to substantially and inevitably rework the moment agentic capabilities arrived.


Designing polished screens on top of a workflow we knew was going to transform inefficient at best. We were solving the wrong problem.


So we paused. We assessed. We pivoted.

04. The solution

First, a new service blueprint. My partner and I went back through the full workflow and systematically identified every step a system could own: 

  • Running calculations, check

  • Detecting anomalies, check

  • Aggregating decision history, check

  • Surfacing relevant context, check 

The opportunities were significant, but we were still mapping against the same old persona-based swim lanes. So we restructured the entire service blueprint around three swim lanes:


100% Human: Decisions, oversight, and sign-off that must stay with the specialist.
100% Agent: Process steps the system can own end-to-end, without requiring human input.
Human-Agent Collaboration: The interaction layer where the agent surfaces its work, explains its reasoning, and the human decides what to do with it.


This framing mattered greatly in how we approached the resulting UI patterns, and it also lended us a story for talking about automation with our project partners. We didn’t say "the system does this now," but: here's where you stay in control.


To make it feel real, I built a UI prototype showing what this workflow could look like end-to-end: a data entry triggers the agent, which runs checks, surfaces an anomaly, decodes it, identifies a likely fix, and presents the specialist with a plain-language summary: here’s what happened, why, and what I think you should do about it. A process that currently takes days could be handled in two clicks (exploding head emoji).


When I showed it to one of the analysts we'd been working with, his response was immediate: "As a user, I want this." Literally the most satisfying thing to hear as a designer.

05. Results & Reflection

The agentic framework is still working its way through leadership approval. (Shoutout Governance & Compliance!) How agentic processes get implemented will take time.


What's already shifted is the team's frame of reference. The question is no longer whether to pursue an agentic approach. It's how. Which agents do we need? What are their responsibilities? How do we define and measure success? How do we build the guardrails?


The most important design decision on this project wasn't a UI choice. It was deciding to stop, reassess, and reframe, even though it meant disrupting momentum. Knowing when to pause is a design skill that doesn't show up in any deliverable, but it's often where the most value gets created.


The resistance we encountered from users worried about their jobs and from an organization still figuring out what it believed about modernization was all data for us to take note of. Designing for resistance rather than around it was what made the pivot land.