easy to work with
attention to detail
great teammate
user centered
UX/UI | Product Design
Riverly
📋 Project Overview
My Role: Lead Product Designer & Product Strategy
Timeline: 30 days (May-June 2025)
Team: Individual contributor (with AI development tools)
Platform: Web application with AI integration
Outcome: Launched product serving 150+ users, 4.9/5 rating, submitted to $1M+ hackathon
🎯 The Challenge
close
75% of professionals experience anxiety during challenging workplace conversations, leading to avoided difficult discussions, missed career opportunities, and escalated team conflicts. Existing solutions (generic chatbots, presentation coaches) don't address the specific context and nuances of workplace communication.
Design Challenge
How might we leverage AI to help professionals navigate workplace conversations with confidence while maintaining authenticity and building genuine communication skills
Success Metrics
User Engagement: Time spent in coaching sessions
Effectiveness: User-reported confidence improvements
Product-Market Fit: User retention and referral rates
Technical Success: AI response accuracy and voice quality
🎯 The Challenge
open
🎯 The Challenge
open
📈 Results & Impact
open
📈 Results & Impact
open
📈 Results & Impact
open
✨ Design Process Reflection
open
✨ Design Process Reflection
open
✨ Design Process Reflection
What Worked Well
User research depth prevented building wrong features
Technical constraints research enabled realistic scope setting
Iterative testing caught major usability issues early
Cross-functional skills allowed rapid prototyping and validation
What I Would Do Differently
Earlier voice interface testing - discovered user hesitation late in process
More diverse user research - limited to my professional network initially
Performance testing earlier - API response time issues emerged during user testing
Accessibility considerations - retrofitted rather than designed-in from start
Skills Developed
AI product design - learned to design for probabilistic rather than deterministic systems
Voice interface design - fundamentally different from visual interface patterns
Technical collaboration - working within API constraints and capabilities
Full-stack thinking - considering backend implications of design decisions
🔍 Research & Discovery
open
🔍 Research & Discovery
open
🔍 Research & Discovery
open
💡 Design Strategy
open
💡 Design Strategy
open
💡 Design Strategy
open
🌱 Design Process
open
🌱 Design Process
open
🌱 Design Process
User Journey Mapping
Core User Flow: "Practice Difficult Feedback"
Entry: User has upcoming 1:1 with underperforming team member
Scenario Selection: Chooses "Constructive Feedback" template
Context Input: Provides specific situation details
AI Setup: System generates personalized coaching prompts
Voice Practice: User practices response, AI provides real-time guidance
Analysis: Post-session feedback on tone, clarity, empathy
Iteration: User refines approach based on AI recommendations
Confidence Check: System assesses readiness for real conversation
Design System & Visual Identity
Brand Strategy
Visual Metaphor: Flowing water representing smooth communication currents
Emotional Tone: Calming yet confident, professional yet approachable
Design Language: Glassmorphic elements suggesting transparency and clarity
Component Library
Glass Cards: Primary content containers with subtle transparency
Flowing Animations: Micro-interactions suggesting water movement
Voice Visualizers: Real-time feedback during practice sessions
Progress Indicators: Skill-building journey visualization
⚒️ Technical Collaboration
open
⚒️ Technical Collaboration
open
⚒️ Technical Collaboration
open
📊 Testing & Iteration
open
📊 Testing & Iteration
open
📊 Testing & Iteration
open
🧠 Key Learnings
open
🧠 Key Learnings
open
🧠 Key Learnings
Design Insights
AI UX Requires Different Design Patterns
Traditional form-based interfaces don't work for AI coaching. Users need:
Conversational flows that feel natural
Progressive disclosure for complex AI features
Transparency in how AI reaches recommendations
Fallback options when AI doesn't understand context
Voice Interface Design is Fundamentally Different
Voice coaching required rethinking basic UX assumptions:
Visual feedback during audio interaction is crucial
Error recovery must be voice-based, not visual
Cultural sensitivity in voice selection impacts trust
Ambient noise considerations for quality experience
Trust Building is Product-Critical for AI
Users need to understand and trust AI recommendations:
Show AI reasoning behind every suggestion
Admit limitations when confidence is low
Provide multiple options rather than single "right" answer
User control over AI guidance level
Technical Learnings
API Integration Strategy
Error handling crucial for user experience
Progressive enhancement allows graceful degradation
Response time optimization more important than feature completeness
User data privacy requires thoughtful architecture decisions
AI Prompt Engineering is a Design Skill
Context setting in prompts dramatically improves output quality
Role-specific language makes AI feel more relevant
Constraint setting prevents inappropriate or generic responses
Iteration cycles similar to design iteration process
Business Learnings
User Research Validates Technical Complexity
Without research showing 79% preference for voice coaching, I would have built text-only interface and missed the key differentiator.
MVP Definition Critical for AI Products
AI capabilities can create feature bloat. Focusing on one core use case (voice coaching) led to better execution than trying to solve all communication problems.
Organic Growth Possible with Right Problem-Solution Fit
Zero marketing spend but strong growth indicates genuine market need being addressed.
About
I specialize in creating digital solutions that not only meet user needs with empathy and insight but also drive tangible business results. By combining deep user research with strategic thinking, I develop designs that enhance user satisfaction while directly supporting key business goals.
Contact