Making day-to-day work with an agent clear and trustworthy.
Once an agent existed, people had to actually work with it — and trust what it did. I focused the everyday experience on visibility and control.
Role · Product & UX design — interaction design, prototyping, UI
The problem
The agent did a lot, but its work happened behind the curtain. Users couldn't easily tell what was in progress, what had happened, or how to step in.
So they either over-checked it or quietly stopped trusting it.
Understanding why
Talking with users, the theme was control, not just information. They wanted to feel in the loop at a glance.
And to be able to take over the moment something looked off — without reading a wall of logs.
The solution
A clear activity surface that shows what the agent is doing in human terms, with obvious handoff points to pause, correct, or take over.
Detail is there on demand, not forced on everyone.
Outcome
Placeholder for shareable results — e.g. higher day-to-day engagement and more self-reported trust.
People could let the agent run and only look when it mattered.
Push toward proactive nudges — the agent surfacing the few moments that genuinely need a human, so people can trust it to run.
Trust isn't built by showing everything; it's built by making the right moment to step in obvious.
Privacy note — screens are recreated with dummy data and details simplified; the real product evolves with business decisions and user needs.