top of page
Enhancing User
Support with
Intelligent Assistance
Leveraging AI to simplify creation, support users instantly, and shape a more efficient, scalable yearbook experience for
12+ Million Daily Users.
2025
6 Months
Senior Product Designer
My Role
Strategy Lead
UI/UX Design Lead
Technical Framing
User Research Lead
65%
Reduction in Tickets
Decrease in routine support inquiries
during peak season.
2.5x
Faster Resolution Time
Average time to resolve user issues
dropped significantly.
92%
User Satisfaction
Decrease in routine support inquiries
during peak season.
95%
Positive Feedback
Positive feedback on the new AI
assistance feature.
The Context
This project explores integrating an AI-powered virtual assistant into the TreeRing Yearbook platform. The goal was to simplify user support, improve operational efficiency, and reduce rising support costs during the busy yearbook season.
I worked as the Senior Product Designer over a six-month sprint, leading the strategy, design, and technical framing to create a scalable solution aligned with critical business goals.
Why It Matters
The growing volume of support requests was degrading both the platform experience and business margins. Without a more efficient model, users would remain frustrated and support costs would skyrocket. An AI assistant offered a sustainable path to reduce load and modernize the support layer.
The Problem
Repetitive Inquiries
Users submitted thousands of duplicate questions because platform information was buried and difficult to find.
Overloaded Support Teams
Customer support was drowning in routine tickets, forcing expensiveseasonal staffing increases.
Fragmented Self-Service
No unified help experience existed, leading to high user frustration and abandonment of tasks.
Top Customer Support Inquiries
Discovery Phase
Mapping Intelligence Around User Needs
Before pixels, we needed to understand the logic. I mapped out complex user scenarios to
determine where AI intervention would be most effective versus where human support
was still necessary.
Competitive Analysis & Research
We utilized Mural to aggregate competitive patterns from leading SaaS
support tools. The analysis revealed that users prefer "context-aware"
suggestions over generic search bars.
40+
6
25+
65%
relied on daily customer support
34%
54%
User
interviews
relied On support
daily
unaware of knowledge articles
12
predictive assistance use cases defined
30+
core user segments identified
journey friction points mapped
aI automation opportunities mapped
User Scenario Mapping
I created this document to align product, engineering, and support around a scalable AI assistance vision, defining user flows, system behaviors, and operational impact while bringing strategic clarity and measurable experience value to the team.
Team & Product Impact
Drove cross-functional AI product alignment
Operationalized scalable support experience strategy
Showing the benefits of reducing cost through intelligent automation
AI Assistant + Sentiment Handling Flow
I created this to help the team design AI interactions that feel more human and supportive, ensuring sentiment, tone, and edge cases were thoughtfully considered across the entire experience.
Team & Product Impact
Brought emotional intelligence into AI conversations
Created consistency across automated support experiencesReduced friction through sentiment-aware response design
The Solution
Building an Impactful Agentic AI Assistant
We designed a non-intrusive, floating assistant that understands the page context. If a user is on the "Layout" page, the AI proactively suggests layout-related tips.
Conversational UI
A context-aware assistant embedded in the editor that delivers real-time, intent-driven guidance to help users move faster and with more confidence.
Natural Language Processing input
Embedded video tutorials
Quick-action suggestion chips
Mobile Accesibility
A mobile-optimized assistant experience that ensures users can access guidance, support, and editing help from any device.
Mobile conversational interface
Embedded responsive tutorials
Tap-friendly quick actionss
Scalable Yearbook Solution
A scalable AI support layer designed to grow with the platform, reducing support dependency while expanding guidance coverage.
Scalable AI knowledge base
Continuously trained models
Automated workflow support
Technical Documentation
Designing System Behaviors Through the User
These documents helped move the product forward by giving teams clarity, supporting engineering build decisions, aligning business and go to market partners, and creating space for the right conversations at the right moments.
System Flow Diagram
Why
Aligned teams on AI vision
Impact
Reduced ambiguity across workflows
Benefit
Enabled scalable product execution
Contextual Layer Mapping
Why
Humanized AI through context
Impact
Reduced confusion during workflows
Benefit
Increased user trust and confidence
Response Decision Logic Tree
Why
Defined AI interaction pathways
Impact
Improved response accuracy and relevance
BenefitEnhanced conversational user experience
Emotional Intelligence Layer
Why
Designed emotionally aware AI interactions
Impact
Reduced user frustration during support
Benefit
Made AI feel more human
Design
Aligning Interface Design With Behavior Logic
We designed a non-intrusive, floating assistant that understands the page context. If a user is on the "Layout" page, the AI proactively suggests layout-related tips.
Visual Design
The visual design established a friendly, approachable AI visual language through soft gradients, rounded containers, and conversational UI styling, supported by a modular component system spanning chat, cards, alerts, and inputs to ensure scalability and consistency. Strong color hierarchy and iconography differentiated system states, guidance, and user actions, while the assistant’s non-intrusive placement balanced visibility without disrupting core creation workflows.
Impact
This visual system made the assistant feel intuitive and approachable, helping users build trust and engage with support more confidently
Interaction Patterns
Interaction patterns were rooted in a conversational support model, enabling users to engage naturally through chat-based guidance. This was enhanced by quick-action chips that accelerated common tasks, embedded tutorials that supported in-flow learning, and expandable assistant states that allowed users to transition from lightweight prompts to deeper support experiences seamlessly.
Impact
Interaction patterns reduced effort and decision fatigue, enabling users to complete tasks faster through guided, conversational support.
Contextual Triggers
Contextual triggers ensured the assistant delivered help at the right moment, surfacing proactive guidance based on page context, workflow stage, and behavioral signals. Support prompts activated during complex or multi step actions, with escalation logic designed to respond to repeated errors, inactivity, or friction keeping assistance timely, relevant, and experience driven.
Impact
Contextual triggers ensured users received the right help at the right time, minimizing friction and preventing workflow disruptions.
Scalable Expierence
One of my core responsibilities was ensuring the experience scaled over time, gradually rolling out trusted, dependable iterations while continuously training the AI model to better support and enhance the yearbook creation experience.
V1: Knowledge Support
Virtual Assistant answers questions through AI integrated tools and pulling from knowledge base articles.
V2: Hands off Approach
The virtual assistant can handle tasks like changing themes or cropping photos, creating a more hands-off and effortless experience for the user.
V2: Hands off Approach
The virtual assistant can handle tasks like changing themes or cropping photos, creating a more hands-off and effortless experience for the user.
bottom of page