Personal experiment exploring how conversational AI can solve real-world design problems through natural language collaboration
Product Designer (solo project) - Applied design thinking principles to guide AI development through iterative conversation
2 days on and off of focused iteration, followed by ongoing refinement and real-world testing
📌 Overview
A product designer's journey from sticky notes to a custom puppy tracker.
🐩 The Problem
As a product designer, I know good UX when I see it. But tracking my Bichon puppy Sarang's bathroom habits? That was pure chaos—sticky notes everywhere, missed patterns, zero insights. I needed a custom solution, but didn't want to spend weeks learning to code.
💡 The Experiment: Design Thinking Meets AI
Instead of wireframes and prototypes, I decided to design with words. Could I use my product thinking skills to guide AI into building exactly what I needed?
My Design Process with AI:
1. Function First → Core Features
"I need to track my bichon's pee and poop times, with consistency data, that works on my phone"
2. Real Data → Import Historical Logs
"I have months of handwritten notes. Can you help me import them? Here's what my logs look like..." [Shared actual messy data format from my iPhone]
3. User Flow → Progressive Logic
"The logging should be a flow—choose type, then consistency, then location, with a record button at the end"
4. Polish → Design System
"Make it look like Airbnb—clean, minimal, whites and grays only"
🚀 What Happened
Through iterative conversation with AI:
Version 1: Getting Something Working
- Started basic: Got a working prototype quickly with colorful gradients and immediate logging
- All options visible at once—functional but overwhelming interface
- Simple tabs for navigation, basic button layout
Version 30: Smart User Experience
- Iterated UX: Refined the user flow through conversation
- Introduced progressive disclosure with numbered steps and conditional logic
- Disabled states and guided process—much better organization
- Still maintaining the colorful aesthetic while improving usability
Version 55: Major Redesign
- Applied design thinking: Used my product skills to guide AI toward cleaner aesthetics
- Started removing excessive colors and implementing more structured layout
- Better typography and spacing decisions
- Transitioning toward professional appearance
Final Version: Professional Polish
- Imported real data: Parsed months of handwritten logs automatically
- Polished to perfection: Achieved Airbnb-quality design with monochromatic palette
- Clean typography, subtle borders, personalized with Sarang's icon
- Export/import functionality and family-ready features
The evolution shows how conversational iteration can transform a basic functional prototype into a professional-grade application through design thinking and continuous refinement.
🎯 What I Discovered
Product Thinking Still Matters
- AI doesn't replace design thinking—it amplifies it. My background in user experience was crucial for:
- Defining the right user flow
- Prioritizing features that actually matter
- Recognizing when something felt "off" in the interaction
Conversational Design is Real Design
- Writing prompts is like writing user stories. The better I articulated the experience I wanted, the better the result:
- ❌ Vague: "Make it better"
- ✅ Specific: "Make consistency options disabled when pee is selected"
Iteration Speed is Game-Changing
- Traditional design process: Sketch → Wireframe → Prototype → Code → Test
- AI design process: Describe → Build → Test → Refine (seamlessly and rapidly)
Focus on Outcomes, Not Implementation
- I described what I wanted the user to experience, not how to code it. AI handled the technical implementation while I focused on the product vision.
🏆 The Result
A professional web app that:
- Looks like a commercial product (Airbnb-quality design)
- Works perfectly on mobile (responsive, touch-friendly)
- Handles real data (imported 90+ historical entries)
- Provides insights (daily patterns, monthly trends)
Most importantly: It actually solved my real problem. My Bichon puppy completed potty training🎉
🔮 What's Next
Right now, this web app only works locally on each device. My next exploration is making it a truly connected experience:
- Cross-device sync: Link mobile and desktop versions
- Family sharing: Let multiple family members contribute to my puppy's tracking
- Real-time collaboration: Everyone sees the same data, updated live
The goal is transforming this personal solution into a family-wide tool that makes pet care a shared responsibility.
🌟 Final Thoughts
This experiment showed me that AI can be an incredibly powerful design tool when you know how to collaborate with it. The key wasn't learning to code, it was learning to communicate my product vision clearly and iterate based on real needs.
For me, this changes everything about how I approach personal projects. Instead of compromising on custom solutions or spending months learning new skills, I can focus on what I do best, understanding user needs and designing great experiences while AI handles the technical implementation.
It's not about replacing human creativity, but amplifying it. AI became my development partner, and my design thinking became the guide.