Agentic Experience (AX)
Agentic Experience (AX): patterns for shipping human–AI collaboration with orchestration, guardrails, explainability, and adoption—without silent failure.
Defined terms I use across design governance, agentic experience, trust, and scalable delivery. Each entry is short, operational, and linked back to evidence and case work.
Agentic Experience (AX): patterns for shipping human–AI collaboration with orchestration, guardrails, explainability, and adoption—without silent failure.
Operational truth: design based on what actually happens in operations—states, exceptions, handoffs, and constraints—so services ship without support debt.
Trust cues: design signals that make risk, intent, and control legible—so users understand what a system is doing and why, especially in high-stakes contexts.
Auditability UX: making decisions, actions, and states traceable and usable for operators, compliance, and investigations—without turning the product into a dashboard.
Recoverability: designing flows and systems so failures can be detected, reversed, retried, or safely handed off—without data loss or user harm.
Consent and control: making permissions, purpose, retention, and revocation understandable and actionable—so users can actually govern their data and identity.
Progressive autonomy: a deliberate ladder from assistive to autonomous behavior, with guardrails, oversight, and reversibility at each step.
Contestability: designing the ability to challenge, correct, and appeal system decisions—so trust survives mistakes and power remains accountable.
Decision artefacts: lightweight documents that capture context, constraints, tradeoffs, and ownership—so teams move faster with fewer reversals.
Governance that ships: operating rhythms, decision rights, and quality gates that increase delivery speed and coherence—without process theatre.
Common core + local layers: a scaling strategy that keeps global coherence while allowing local adaptation—without fragmentation.
State model: a shared definition of system states and transitions that makes complex journeys governable, recoverable, and auditable.
Evidence surfaces: UX patterns that show the basis of a recommendation or decision—inputs, constraints, provenance—so outcomes remain explainable and contestable.
Escalation paths: predefined handoff routes for blocked decisions and high-risk uncertainty—so teams don’t hide problems until launch.
Quality gates: lightweight, risk-based checks that protect high-stakes journeys—so teams ship faster with fewer reversals and fewer incidents.
Let’s discuss a leadership role, advisory work, or a complex product challenge.