A self-paced course on the design practice behind AI-powered products: how interfaces should behave when the system is probabilistic, slow, sometimes wrong and increasingly able to act on the user’s behalf. Six modules cover how AI changes the design contract, conversational and prompt-driven interfaces, communicating confidence and uncertainty, human oversight of agents, building an AI design system with reusable components and tokens, and evaluating AI experiences with research and telemetry.

Written for product designers, front-end engineers and product managers who are adding generative or agentic features to existing products and need patterns that hold up in production. Tech Skills Library is independent and issues no certification, credential or licence.

Modules

Module One — How AI Changes the Design Contract

Lessons

• Designing for Probabilistic Outputs Instead of Deterministic Screens Preview
• Mapping User Intent to Model Capability Boundaries
• Setting Expectations Before the First Prompt

Module Two — Conversational and Prompt-Driven Interfaces

Lessons

• When a Chat Interface Is the Wrong Choice
• Structuring Prompt Inputs With Scaffolds and Constraints
• Multi-Turn Context and Conversation Memory Controls

Module Three — Trust and Confidence Communication

Lessons

• Confidence Indicators That Do Not Mislead
• Citations and Provenance in Generated Answers
• Designing Honest Refusals and Graceful Fallbacks

Module Four — Human Oversight of Autonomous Agents

Lessons

• Levels of Agent Autonomy and Where to Put the Approval Gate
• Plan Previews and Step-by-Step Agent Transparency
• Interruption Undo and Rollback Patterns for Agent Actions

Module Five — Building an AI Design System

Lessons

• Component Inventory for Generative and Agentic Features
• Design Tokens for Uncertainty Latency and Machine Authorship
• Content Design Guidelines for Machine-Generated Text

Module Six — Evaluating and Iterating AI Experiences

Lessons

• Usability Research Methods for Non-Deterministic Features
• Feedback Capture and Rating Controls That Produce Signal
• Product Telemetry and Quality Dashboards for AI Features