How I made a complex enterprise data synchronization workflow easier to understand, configure, and recover from.
2–3 hours → under 15 minutes
Setup time
At a glance
Connection and status visibility
Traceable
Failure feedback with timing context
The public version focuses on the design reasoning and outcome, while omitting proprietary screens, internal identifiers, and implementation details.
Challenge
Sustainability teams were reconciling information between an internal platform and an external benchmarking service. Before they could act, they had to understand ownership, direction, permissions, status, and what to do when something failed.
Before
After
Approach
I combined competitive analysis, technical feasibility assessment, conversations with service users, wireframing, prototyping, and usability testing. Each activity answered a different question: what do people expect to see first, which terms create hesitation, and which feedback helps them recover without outside support?
Research
I mapped the moments where people needed orientation, interpretation, or reassurance before taking action.
Feasibility
API and workflow constraints shaped what could be surfaced immediately and what belonged one layer deeper.
Validation
Usability testing exposed small wording and feedback gaps that made the workflow feel less trustworthy than it was.
Design Direction
01 / Connection
Make the linked account, ownership, and latest activity visible before a user opens a detail view.
02 / Direction
Use direct language and clear status cues so the relationship between systems is understandable at a glance.
03 / Recovery
Pair a visible failure state with plain-language context and timing so the next step is easier to find.
Outcome
The final direction brought connection status, data direction, and recovery feedback into one clearer mental model. It gave teams a faster path through setup while making the system's boundaries and failure states more legible.
Users can understand the state of the relationship before opening configuration.
Visual status cues reduce the need to read every row to find what needs attention.
A failure carries enough explanation and timing to support the next conversation or action.
Reflection
A technically accurate label can still fail if it does not match the user's mental model. Naming deserves the same care as the interaction around it.
A status becomes more useful when it explains what changed, when it changed, and what the person can do next.
Working within existing systems pushed me to prioritize the moments that most improved confidence instead of adding complexity for its own sake.