Agentic User Experience (AUX)
Pioneering Agentic UX — where AI agents meet human-centred design.
I design autonomous agent workflows that help people express intent, define boundaries, monitor progress, and stay in control — without drowning in system noise.
What is Agentic UX?
Agentic UX (AUX) is the design discipline for products where AI agents act on behalf of users — translating high-level goals into multi-step execution, making decisions within defined boundaries, and collaborating with humans through transparent handoffs and approvals.
Unlike traditional UX where every interaction is deterministic, AUX deals with autonomy, confidence, fallback paths, and trust calibration between humans and AI agents.
The AUX Framework
Intent capture
Help users express goals clearly — including constraints, preferences, risk tolerance, and success criteria — so the agent can act with purpose.
Autonomy boundaries
Define what the agent can do independently, what needs human review, and when a person must approve, override, or stop execution.
Transparent execution
Show progress, decisions, assumptions, confidence levels, and next steps without overwhelming users with internal system noise.
Human-in-the-loop
Design smooth handoffs for approval, escalation, correction, and audit at key decision points in agentic workflows.
Recoverable errors
When agents make mistakes, users need clear paths to correct, retry, rephrase, or escalate — not dead ends or opaque failures.
Why enterprise AI needs AUX
Enterprise AI adoption depends on trust. Users must understand what the system is doing, why it chose a particular action, and how to intervene when things go wrong. AUX provides the design patterns for:
- Research-first validation of AI assumptions before building agentic features.
- Interfaces that show status, sources, confidence, and next actions — not just raw model output.
- Clear controls for approve, reject, regenerate, edit, escalate, and audit agent decisions.
- Accessibility and consistency across every AI surface through design system integration.
How my background maps to AUX
At Ericsson, I led UX for GenAI initiatives where I designed interfaces that convert voice and text prompts into production-ready UI concepts. My telecom orchestration experience — where systems translate high-level goals into technical execution with human oversight — is directly relevant to AI agent products.
- GenAI workflows that convert prompts and Figma wireframes into production UIs.
- Telecom orchestration experiences where automation and human operators collaborate seamlessly.
- Design systems that keep AI-assisted interfaces consistent, accessible, and governable.
- Usability research to validate whether users trust and understand AI-mediated flows.
Explore related profiles
AI UX Designer
GenAI interaction design, prompt-driven flows, and AI-ready design systems.
UX Research
How I validate complex workflows with real enterprise users.
Senior UX Designer Ireland
Senior UX leader available in Ireland and remote.
Building agentic products?
I am available for AUX strategy, GenAI UX design, and enterprise AI product leadership roles.