Human-AI Product Designer

Designing for delegating, interrupting, and reviewing autonomous agents.

Multi-turn dialogue design, intent modeling, and repair flows.

Confidence indicators, source attribution, and model limitation disclosure.

Approval gates, escalation paths, and override affordances.

Designing prompt composition surfaces, suggestions, and templates.

Graceful degradation, error states, and recovery flows for AI features.

Voice + touch + visual coordination across modalities.

User mental model alignment, expectation setting, and over-trust mitigation.

Screen-reader-friendly AI output, cognitive load, and reading-level tuning.

Character design for AI assistants, consistency, and contextual adaptation.

Task success, satisfaction, and reliance metrics for AI-assisted workflows.

Progressive disclosure, first-run experience, and capability scaffolding.

Consent, dark-pattern avoidance, and dignity-preserving automation.