AI Patterns
Battle-tested design patterns for building AI-powered products — covering onboarding, input, output, trust, control, refinement, memory, and spatial orchestration. Pro members get the full implementation guide for every pattern.
Discovery & Onboarding
- Expectation-Setting Disclosure
Clarify AI involvement, limits, and what users should verify before relying on outputs.
High UX ImpactLow - Suggested Prompts
Provide pre-written prompts or starter actions that help users begin quickly.
High UX ImpactLow - Proactive Suggestions
Surface timely AI recommendations or next-step actions before the user asks.
High UX ImpactHigh - Icons / AI Presence Signals
Use visible badges, icons, and markers so users can quickly notice AI presence or activity.
Medium UX ImpactLow
Input & Intent Expression
- Open Prompt InputPro
Offer a flexible natural-language input so users can ask in their own words.
High UX ImpactLow - Structured PromptingPro
Guide input with fields, templates, or scaffolds that make prompts clearer and more reliable.
High UX ImpactMedium - Model and Mode TuningPro
Let users adjust the AI’s mode, model, or behavior when it meaningfully affects results.
Medium UX ImpactMedium - Inline Prompt AssistancePro
Help users compose better prompts in the moment with inline guidance and suggestions.
High UX ImpactMedium - Searching & FilteringPro
Let users search naturally and refine results with filters, facets, or follow-up constraints.
High UX ImpactMedium - Multimodal InputPro
Support text, voice, image, and file inputs alone or in combination.
High UX ImpactHigh - Context GroundingPro
Ground the model in relevant files, sources, or workspace context instead of guesswork.
High UX ImpactHigh
Output & Response
- Alternative GenerationPro
Show multiple candidate outputs so users can compare and choose the best one.
High UX ImpactMedium - Structured Response FormattingPro
Return outputs in clear, predictable formats that are easy to scan and reuse.
High UX ImpactMedium - Progressive Result RenderingPro
Reveal results as they are generated instead of making users wait for the full output.
High UX ImpactMedium - Actionable OutputPro
Present results in forms users can immediately apply, edit, approve, reject, or send onward.
High UX ImpactMedium
Transparency & Trust
- Process VisibilityPro
Make the system’s current activity, progress, and tool use legible while it works.
Medium UX ImpactMedium - Source and Provenance TransparencyPro
Show where information came from and how content was generated or transformed.
High UX ImpactMedium - Confidence and Uncertainty CommunicationPro
Signal certainty, uncertainty, and when extra verification is needed.
High UX ImpactMedium
Control & Safety
- Adjustable AutomationPro
Give users control over how much work the AI performs on their behalf.
High UX ImpactHigh - Human Review and TakeoverPro
Keep humans in the loop with clear approval, editing, and takeover points.
High UX ImpactHigh - Failure RecoveryPro
Help users recover cleanly when the AI is wrong, blocked, interrupted, or incomplete.
High UX ImpactMedium