Product narrative · 9 keyword pages
Goal-based interaction is a paradigm where the user expresses what they want to achieve (the outcome) rather than specifying the steps to get there. Instead of clicking through five screens to book a flight, the user says 'get me to Berlin next Tuesday, cheape...
Intent-driven UX designs interfaces around detecting and fulfilling user intent rather than presenting fixed menus of options. The system interprets what the user is trying to do — through natural language, behaviour patterns, or contextual signals — and adapt...
Proactive UX describes interfaces that take initiative — surfacing information, suggesting actions, or starting tasks before the user explicitly asks. A proactive agent might flag that a customer's contract is about to expire, draft a renewal email, and schedu...
Asynchronous interaction design handles the UX of tasks that happen over time rather than in real-time. When an agent takes 30 minutes to research a market, the user shouldn't be staring at a loading spinner. Async design covers delegation patterns (hand off a...
Human-in-the-loop (HITL) systems require human approval or input at specific points in an automated workflow. The human is a mandatory checkpoint — the agent cannot proceed without sign-off. HITL design covers approval interfaces (making decisions fast and inf...
Human-on-the-loop (HOTL) differs from HITL in that the agent can proceed autonomously, but a human monitors and can intervene. The human is a supervisor, not a gatekeeper. HOTL design focuses on monitoring dashboards, exception alerting (flagging only anomalie...
Adaptive interfaces change their structure, content, or behaviour based on the user's context, expertise, preferences, or task state. For agentic systems, this means the interface might show simplified views for new users and detailed agent-activity logs for p...
Feedback loops in AI systems are the mechanisms by which user actions, corrections, and preferences flow back to improve agent behaviour. This includes explicit feedback (thumbs up/down, corrections, ratings), implicit feedback (which suggestions users accept...
Collaborative AI workflows are processes where humans and agents work together on shared tasks, each contributing their strengths. The agent handles data gathering, pattern recognition, and repetitive execution; the human provides judgment, creativity, and con...