Course Objective & Scope
If users watching a screen-acting or voice agent wonder 'What is it doing right now, and how do I stop it if it clicks the wrong button?', they abandon the tool; yet popping up a confirmation modal on every step causes confirmation blindness. This programme covers the design of confirmation, visibility and control mechanisms in Agent Experience (AX).
Participants design risk-aware confirmation cards, spoken verification protocols, live supervision views and takeover/handback flows, producing a human-confirmed agent prototype for a scenario adapted to their organisation.
Who Should Attend?
- Product managers and UX/UI designers building AI and agent products
- Frontend and full-stack agent developers
- Digital channel safety and customer experience teams
- Prerequisite: basic digital product design or application development background
Key Takeaways
- Map agent actions into silent execution, toast notification or explicit confirmation tiers by impact and reversibility
- Design action preview and diff cards that show users exactly what will change at a glance
- Build live supervision, panic-stop and seamless human takeover/handback experiences for UI agents
- Write explicit and implicit spoken confirmation and undo patterns for voice assistants
- Define metrics to track confirmation fatigue and accidental approval rates
Curriculum
Control and Visibility Principles in Agent Experience (AX)
- Showing users what the agent is currently planning and about to execute
- Tiered permission models that prevent confirmation fatigue
Visual and Structured Confirmation Patterns
- Pre-action summary cards, before/after diff views and inline editable parameters
- Designing deliberate friction for high-stakes or irreversible actions
Live Supervision and Takeover in Computer Use
- Picture-in-Picture virtual screen, cursor highlights and step timelines
- Handing control to the user for CAPTCHAs, MFA or ambiguity and handing back cleanly
Spoken Confirmation and Undo in Voice Interfaces
- Implicit confirmation ('Booking that for Monday at 10...') versus explicit consent
- Grace-period cancellation windows and state rollback flows
UI Safety Boundaries and Prototype Workshop
- Defending against deceptive UI elements, hidden text and clickjacking on external sites
- Human-confirmed prototype and UI safety checklist
Other Courses in This Category
-
Testing voice, multimodal and UI-acting agents for user experience, trust and safety before launch: user trust calibration, adversarial scenarios, frustration signal detection and release criteria.
- Live Course
- 5 modules
-
Multimodal model architectures that unify vision, audio and text representations: ViT/SigLIP and audio encoders, cross-modal projection/fusion layers, and LoRA fine-tuning of open-weight VLMs (Qwen-VL, Llama, Pixtral).
- Live Course
- 5 modules
-
Building visual inspection and analysis applications with Vision-Language Models (VLMs) in manufacturing, retail, insurance and field operations: visual grounding, multi-image comparison and hybrid CV+VLM pipelines.
- Live Course
- 5 modules
-
End-to-end Speech-to-Speech (S2S) models and autonomous voice agents: acoustic emotion and prosody perception, voice tool calling, multi-agent call handoffs and production architecture with Pipecat and LiveKit.
- Live Course
- 5 modules
-
Streaming infrastructure engineering for real-time AI applications: WebRTC, WebSockets and SSE protocols, OpenAI Realtime and Gemini Live API integration, audio/video buffer management and horizontal scaling.
- Live Course
- 5 modules
-
Human-like turn-taking and millisecond latency engineering in voice AI: acoustic VAD, semantic endpointing, false-interruption suppression, TTFT/TTFB optimisation and speculative execution.
- Live Course
- 5 modules
Get in touch
