top of page
Designing AI-Powered Photo Curation & Layouts
Automating photo layout generation to reduce manual design effort while empowering user creativity.
2025
8 Months
Senior Product Designer

Impact At Scale
Our measurable improvements in user experience, efficiency, and business
outcomes
65%
Reduction in Tickets
Decrease in routine support inquiries
during peak season.
50%
Time Reduction
Reduction in manual page-building time
↑ 65%
Completion Rates
Layout completion rates increased
↑ 78%
User Confidence
Average time to resolve user issues
dropped significantly.
65%
Reduction in Tickets
Confidence in design creation
92%
User Satisfaction
Decrease in routine support inquiries
during peak season.
↓ 42%
Support Requests
Reduced layout editing support tickets
Business Outcomes

Accelerated Book Creation Timelines
Users complete projects faster with AI-assisted layouts

Reduced Design Friction
Non-designers empowered to create professional layouts

Increased Platform Retention
Higher satisfaction drives repeat usage year-over-year

Enabled Scalable AI Feature Expansion
Foundation for future intelligent automation features
The Challenge
Creating yearbook layouts was time-intensive and overwhelming for many users,
particularly those without design experience. Users struggled with blank-canvas
paralysis, layout decisions, and time constraints, often delaying book completion.
User Pain Points
Didn't Know Where to Start
Blank canvas paralysis prevented users from beginning projects
Spent Hours Arranging Photos
Manual photo placement consumed excessive time and energy
Felt Unsure About Design Quality
Lack of design confidence led to project abandonment

Business Impact

Low Feature Adoption
Complex tools remained underutilized by target users

Delayed Project Completion
Extended timelines affected revenue and customer satisfaction

Increased Support Tickets
Overwhelmed support teams with layout-related questions
My Role as Senior UX / Lead
I led the experience strategy and design execution for AI-powered layout
generation, partnering with product and engineering to define automation logic,
layout frameworks, and user control systems.

Strategy & Vision
Defined AI layout generation strategy
Designed layout logic frameworks

Design Execution
Built user customization controls
Led prototyping and testing

Collaboration
Partnered with engineering teams
Balanced automation with creative flexibility
Research
Understanding Behaviors
Deep research into how users interact with layout tools and what drives successful yearbook creation.
Research Methods

User Interviews
35+ in-depth conversations with yearbook creators

Layout Usage Analytics
Analyzed cross compietitor layout creation sessions

Session Recordings
Observed real user behavior patterns
Layout Usage Analytics
Analyzed 10K+ layout creation sessions
Layout Usage Analytics
Analyzed 10K+ layout creation sessions
Layout Usage Analytics
Analyzed 10K+ layout creation sessions

Support Ticket Reviews
Identified common pain points and friction
Key Insights
1
Wanted Guidance, Not Full Automation
Users desired intelligent suggestions while maintaining creative control
Creative Control Was Critical
Ability to override and customize layouts was non-negotiable
Layout Suggestions Reduced Anxiety
Starting point eliminated blank canvas paralysis
Time Savings Drove Satisfaction
Faster completion directly correlated with positive sentiment
3
4
2
AI-Assisted Research Process
Leveraged AI tools to accelerate research synthesis and uncover deeper insights

ChatGPT and Gemini For Research Synthesis
Market Research Analysis
Processed compeitor analysis data and industry reports to idenftify market gaps and opprotunities in AI-assisted design tools
Quantifying Research Insights
Synthesized interview transcripts and responses to extract quantifiable patterns and statistical signigicance.
Insight Generation
Transformed raw user feedback into actionable design principles and feature requirements.

Figma Make for Layout Exploration
Rapid Prototyping
Generated multiple layout variations quickly to test different grid systems and visual hierarchies.
Design System Scaling
Accelerated component creation and maintained consistency across layout templates.
Iteration Speed
Tested design concepts 3x faster with AI-assisted layout generation and auto layout features.
System Design Thinking
Designing the Layout Intelligence System
We designed an AI-assisted layout engine that generated structured,
aesthetically balanced page designs based on photo quantity, orientation, and
hierarchy rules.

Building User Trust Though Logic
Intelligent algorithm considers multiple factors to create balanced, professional layouts

Photo quantity analysis

Orientation detection (portrait/landscape)

Visual hierarchy rules

Compositional balance scoring
Creating Alignment With Strategy

Understanding the concerns, finding an opprotunity for a solution while creating alingment amongst the teams on strategy.

Drafting The Empty/Fallback State Diagrams
Understanding the concerns, finding an opprotunity for a solution while creating alingment amongst the teams on strategy.

AI API Curation Experience Framework

Created to translate technical API logic into transparent user-facing photo selection rules, ensuring trust and clarity within automated layouts.
My System Designe Approach
This project elevated my role from product designer to system architect, defining the intelligence layer that powers layout generation while
ensuring human creativity remains central to the experience.

Logic Definition
Aligned JSON specifications with AI curation triggers to clarify photo quality and selection guidelines within the experience.

Parameter Tuning
Established guardrail logic and fallback state interactions to protect quality and preserve experience continuity.

Scalable Architecture
Designed a scalable foundation to enable future AI enhancements and evolving creative tools within the yearbook platform.
Visual & UI Design
Balancing Automation & Creativity
The core design challenge: provide intelligent assistance without removing the joy of creative expression


Layout Suggestions
AI generates multiple layout options instantly based on uploaded photos. Users can preview different arrangements before committing, eliminating guesswork and providing creative inspiration.
Multiple layout variations generated instantly
Real-time preview before applying
Context-aware suggestions based on page type

Providing Trustable
Insights
Designed an explainability layer where the AI surfaces rationale for excluded photos, enabling informed user review before print and reducing operational risk.


Giving Full Control At All Times
The AI generated layout integrates into the core editor experience, ensuring users retain full creative control and uninterrupted access to all standard design capabilities.


Creating Comfort with Familiarity
Strategically retained established brand patterns and embedded the functionality within the primary editor workflow to maximize adoption, recognition, and user confidence.

Designing Flexible Editing Controls
Micro-interactions and editing workflows that feel intuitive and responsive and trutworthy
Refine AI Selected Photos
The JSON driven system curates optimal photo sets based on defined constraints while intentionally preserving user agency to strengthen trust in AI assisted decisions.

Support Creative Expansion Beyond Automation
After the layout is generated, users can seamlessly adjust photo sizes and refine the page, enabling ongoing creativity while the system functions as an intelligent assistant rather than a rigid automation tool.

System Documentation
Technical Documentation & UI Behavior
Comprehensive documentation defining how the AI layout system should
behave, edge cases, and interaction patterns

Interaction State Definitions
Folder Selection
Users begin by selecting a photo folder, allowing the system to analyze content volume and prepare structured layout recommendations
Intelligent Image Curation
The AI evaluates duplicates, resolution, composition, and quality to surface the strongest images while clearly flagging any print risks.
Guided Layout Selection
Multiple layout variations are generated based on selected images, giving users structured options with full review and control before placement.
Seamless Editor Integration
The selected layout drops directly into the yearbook editor, preserving full editing capabilities so users can refine, rearrange, and expand creatively.

Edge Case Handling
Low Resolution Warning
Images that do not meet print quality thresholds are flagged with clear messaging to prevent blurry or pixelated output.
Duplicate Detection
Visually similar photos are automatically filtered, with transparent feedback to avoid redundancy and improve layout variety.
Orientation Mismatch
Photos that conflict with layout orientation constraints are repositioned or recommended for alternate slots to preserve composition integrity.
Low Photo Count
When too few photos are available, the system adapts layout density or recommends uploading more images to maintain visual balance.
Design Standards
Design System & Visual Language
Established comprehensive design standards to ensure consistency and quality
across all AI-generated layouts

Visual Hierarchy Rules
Photo Size Ratios
Hero
1.5-2x
Secondary
1x
Supporting
0.5-0.75x
Spacing Standards
Minimum gutter:
Recommended gutter:
Page margins:
8px:
16px
24-32px:

Layout Quality Metrics
Visual Balance Score
Measures weight distribution across quadrants
PASS: ≥7/10
White Space Ratio
Ensures breathing room and readability
PASS: 15-30%
Hierarchy Clarity
Focal point identification and size differentiation
PASS: ≥8/10

Grid System Framework
12-Column Base Grid
• Layouts adapt to 12, 6, 4, 3, 2, or 1 column variations
• AI selects grid based on photo count and orientation
• Custom spans allowed for featured content

Accessibility Standards
Minimum touch target size: 44x44px for interactive elements
Alt text required for all photos before layout generationKeyboard navigation support for all editing functions
Screen reader announcements for AI generation status
WCAG AA contrast ratios for all UI text elements
Design Principles

Intelligent Defaults
AI suggests optimal layouts while
allowing full customization

Visual Clarity
Clear hierarchy and breathing
room in every generated layout

Performance
Sub-2s generation time with
instant preview updates

Consistency
Unified design language across all
layout variations
Product Thinking
Key Product Decisions
Strategic decisions that shaped the product direction and balanced competing
priorities

Automation vs Control
Decision: 70/30 split favoring user control
We tested full automation but users felt "locked out." The sweet spot: AI generates initial layouts, but every element remains editable.
"I want suggestions, not decisions made for me."
— User Testing Feedback

Pure randomness created chaotic layouts. We constrained variation to
maintain professional quality while still offering diversity.
Variation Range
Moderate
Layout Randomness Thresholds
Decision: Controlled variation within aesthetic bounds

Grid Rigidity vs Flexibility
Decision: Flexible grid with snap guidelines
Strict grids felt limiting; freeform was too chaotic. Magnetic snap points
provided freedom with guardrails.
Snap-to-grid suggestions (optional)
Free positioning available
Alignment guides on demand

AI Confidence Scoring
Decision: Transparent confidence indicators
We show users when AI is "confident" vs "experimental" in layout
suggestions—building trust through transparency.
High Confidence
Experimental
Live & Launched
In Production Today
The AI-powered layout system is live in production, actively used by thousands
of yearbook creators across schools worldwide.

Launch Milestones

August 2025 - Beta Launch
Rolled out to 50 pilot schools for initial testing and feedback collection

September 2025 - Full Release
Launched to all platform users with 82% feature adoption rate within first month

Platform Integration

Web Platform
Fully integrated into primary web editor with real-time
sync

Mobile App
iOS and Android apps with optimized touch interactions


Road Map Ahead
Weekly ML model retraining based on user selections
AI-powered caption suggestions
Smart photo tagging and face recognition
Theme-aware layout generation
Multi-page spread optimization
Outcomes at Scale
The AI layout system transformed the yearbook creation experience and
delivered measurable business value

User Pain Points

Reduced Manual Design Workload
50% average time savings in layout creation phase

Improved User Satisfaction
4.6/5 average rating on AI layout feature

Empowered Non-Designers
78% increase in confidence creating layouts

Business Impact

Increased Feature Adoption
65% uptake in first 3 months post-launch

Accelerated Book Completion
Projects finished 40% faster on average

Higher Retention Rates
Year-over-year customer retention improved 23%
Evolved from UI design to
architecting intelligent systems
System Thinking
This project expanded my ability to design AI-assisted creative tools—balancing
automation with human expression while aligning business goals with user
empowerment.

Growth as a Design Lead
Cross-Functional Leadership
Evolved from UI design to
architecting intelligent systems

Connected UX decisions to business outcomes
Strategic Vision

bottom of page
