The Role

This is not a traditional UX role. Most design assumes a human drives and software responds. You‘ll design the inverse: autonomous agents act, and humans supervise, intervene, correct, and trust. You‘ll define how staff oversee fleets of agents — what to surface, when to interrupt, how to hand control back and forth, and how to make an agent‘s reasoning legible enough that a human can step in confidently mid-task (live-call barge-in, supervisor consoles, exception queues, confidence-gated approvals). You own the design language for human-agent collaboration in a regulated healthcare environment, where a misread interface has real consequences.

What You‘ll Do

• Design supervision & intervention experiences: live monitoring, escalation, human-in-the-loop approvals, and graceful agent↔operator handoff.
• Create patterns for agent legibility — communicating intent, confidence, state, and rationale...