AI Skill Report Card
Designing AI Powered Apps
AI-Powered App Design
Quick Start15 / 15
App Concept: Smart Expense Tracker
Core AI Feature: Receipt scanning with automatic categorization
Key UX Flow: Photo → AI Processing → Confirm/Edit → Saved
UI Components:
- Camera capture button (primary CTA)
- Processing state with friendly animation
- Editable card layout for AI suggestions
- One-tap confirmation system
Recommendation▾
Add concrete input/output pairs for AI prompting examples instead of just showing prompts
Workflow13 / 15
Progress:
- Discovery: Define user problem and AI solution fit
- AI Strategy: Map AI capabilities to user value (not tech features)
- Core Flow: Design the 3-tap user journey for primary use case
- AI UX Patterns: Handle loading, errors, and confidence levels
- Interface Design: Create clean, gesture-friendly layouts
- Prototype: Build interactive mockup with realistic AI delays
- Testing: Validate with users using wizard-of-oz AI simulation
Recommendation▾
Include specific design templates or wireframe frameworks rather than just principles
Examples18 / 20
Example 1: Fitness Coach App Input: Users want personalized workout plans AI Integration: Computer vision for form correction + NLP for motivation UX Solution:
- Live pose detection with real-time visual feedback
- Conversational AI coach that adapts tone to user progress
- Progressive disclosure: show corrections only when needed
Example 2: Recipe Discovery App Input: Users have ingredients, want meal suggestions AI Integration: Image recognition + preference learning UX Flow:
- Multi-select ingredient camera scan
- AI generates 3 ranked options with confidence scores
- One-tap cooking mode with smart timers
Recommendation▾
Provide more technical implementation details for AI integration patterns
Best Practices
AI UX Principles:
- Show AI confidence levels visually (progress bars, color coding)
- Always provide edit/override options for AI suggestions
- Use loading animations that indicate AI is "thinking"
- Design for AI failures with graceful fallbacks
Interface Design:
- Follow platform guidelines (Material Design/Human Interface)
- Use 8pt grid system for consistent spacing
- Design for thumb-friendly interactions (44px+ touch targets)
- Implement dark mode from start, not as afterthought
AI Prompting in Design:
- Use specific constraints: "Design a button for users with motor impairments"
- Include context: "Create onboarding for elderly users new to AI"
- Request multiple variations: "Show 5 different ways to display AI confidence"
Common Pitfalls
- Don't expose AI complexity to users (no "neural network processing")
- Don't make AI the hero - solve user problems, not showcase technology
- Don't design flows longer than 3 steps for primary actions
- Don't use generic AI terms ("smart," "intelligent") - be specific about value
- Don't skip empty states and error handling for AI features
- Don't assume AI will be 100% accurate - design for iteration and learning