projects Overview
Context
Little Loop is a 0→1 product initiative inspired by my parenting experience and commitment to sustainability. It explores how AI can address everyday challenges for families. I applied AI not only in the product, but also in my design process, with the goal of making exchanges easier and more joyful.
Challenge
As a mom, I often shared preloved children’s items through Buy Nothing and local Facebook groups. The exchanges were meaningful, but the process was frustrating—posting was tedious and coordinating pickups was clunky. For busy parents, these hurdles turned sustainable sharing into a time-consuming chore.
Solution
A caregiver-friendly platform that connects giving and receiving through a smart matching system. AI-assisted listing creation, paired with built-in scheduling and messaging tools, makes sharing and coordination effortless—saving parents’ time while keeping the joy in passing things along.
Contribution
Created as a proof of concept, this project evolved from research and insight gathering to concept design and a live, AI-powered prototype built through vibe-coding. Testing with parents helped validate usability and uncover opportunities that now shape the next iteration.
Validation
2x
Fewer steps in Listing creation
3x
Fewer messages for Pickup Coordination
•
Overall high satisfaction, with parents describing the experience as smooth, intuitive, and time-saving
•
Preferred over Facebook Marketplace and Buy Nothing / local Facebook Groups for being more purposeful and less mentally taxing
Problem Space
Discovery & Insight
To move beyond my own experience, I surveyed parents in local communities to better understand the challenges of exchanging pre-loved children’s items. The responses echoed what I had felt and revealed opportunities for improvement:
What 28 Parents Told Me:
Key Pain Points
•
Pickup coordination (75%)
•
Incomplete listing details (71%)
→ Most critical info to users is pickup location, condition/appearance, age suitability, product specs, and listing status
•
Difficulty finding specific items (54%)
•
Responding to inquiries (46%)
•
Creating listings (39%)
•
Competition (25%)
→ Due to items being claimed quickly or chances of receiving are low
Children's Age Range
•
Most users have children aged 3–5 years (68%) and 0–2 years (43%)
Popular Exchange Items
•
Toys/play equipment (89%)
•
Clothing/accessories (75%)
•
Baby gear/equipment (68%)
Most Used Digital Platforms
•
Buy Nothing Groups (75%)
•
Local Facebook groups (50%)
•
Facebook Marketplace (50%)
•
Craigslist (10%)
Use Cases
•
Primarily giving items away (89%) rather than asking for items (54%)
Key Motivations
•
Giver: Decluttering (100%)
•
Receiver: Saving money (96%)
•
Sustainability (86%)
•
Community support (71%)
Interest in Community Features
•
A sense of connection/trust with other parents is significant
•
Simplicity and speed were preferred, with little enthusiasm shown for gamification features
•
Sustainability-focused rewards could be a meaningful motivator
Design Opportunities
Identifying key design opportunities
•
Bridge the gap between givers and receivers:
Make exchanges more efficient by connecting the right items to the right people.
•
Simplify listing creation while improving quality:
Reduce the mental load for givers and ensure listings include the details receivers care about most.
•
Enhance discovery and navigation:
Personalize content and improve browsing so parents can quickly find what they need.
•
Streamline communication and pickup coordination:
Centralize messages, cut unnecessary steps, and make scheduling faster and easier for both sides.
•
Build a community-centered sharing experience:
Encourage genuine, trust-based, and lasting connections beyond one-time exchanges.
design Principles
Little Loop was designed with the belief that technology should make human connections easier, not replace them. These principles guided how I integrated AI—keeping the experience simple, transparent, responsible, and thoughtful while helping parents share more meaningfully.
Empowerment over automation
AI should assist, not replace. It accelerates repetitive tasks while keeping the emotional fulfillment of human connection intact. Parents stay in control with personalization settings, edit options, and clear feedback mechanisms to help improve the system overtime.
Trust + Transparency
Explain the magic behind AI system’s outputs. Ensure the users understand why certain decisions are made in AI-powered listing creation, listing recommendations, and communication assistance.
Simplicity + Clarity
Prioritize time-saving, intuitive workflows for busy parents. Keep every step easy to follow and every listing accurate and useful.
Context-Aware Design
Draw on contextual cues — such as child’s age, season, and past exchanges — to deliver smarter suggestions, more relevant listings, and seamless coordination.
Concept Development
I mapped select user flows to illustrate the system logic behind core interactions, showing how key moments connect across the giving and receiving journey.
Wireflows spotlight the giver’s experience, highlighting how thoughtful UX decisions simplify listing and coordination:
Prototyping Process
Exploring, building, and learning through vibe-coding and a fluid design process
I experimented with vibe-coding and integrated an AI model to build a live-data prototype that brought early design ideas to life. Along the way, I learned through trial, debugging, and iteration—navigating compromises and trade-offs while gaining hands-on insight into designing adaptively with emerging technology.
Feature Priority
I explored multiple ideas across user flows within the opportunity space. To stay focused, I prioritized designing and prototyping features that addressed the most pressing user pain points and offered the greatest impact:
•
•
•
PRD
I used ChatGPT to quickly draft a PRD outlining the core features and functionalities for the first prototype. It also served as the initial prompt in Bolt to scaffold the project structure and set up the basic environment.
Tech Stack
Researching with the help of a few LLMs, I identified the tools and APIs for building this prototype:
•
Bolt – Built and tested an AI-powered prototype with live data.
•
Supabase - Managed backend data storage and real-time updates.
•
Google Cloud Vision API – Used computer vision to extract product details from uploaded images.
•
Google Gemini API – Powered the AI listing feature, auto-generating item details and category suggestions.
•
Google Maps & Geocoding APIs – Supported location-based search, distance filters, and pickup location mapping.
•
Google Custom Search API – Provided supplemental content and data enrichment.
Implementing these tools, AI models, and API integrations was new territory for me, but I leaned into the learning curve—experimenting, problem-solving, and ultimately building a live prototype that made the concept tangible.
Simplifying Listing
Creation with AI
AI analyzes uploaded images, identifies the item, and suggests key details to simplify listing creation.
•
Maintain a human, trusted tone:
Guide AI-generated descriptions to feel parent-to-parent—casual, warm, and non-commercial.
•
Make key info complete:
Have AI capture core item details (condition, size, age range) and pair them with the user’s pickup location to create a complete listing.
•
Design for trust & transparency:
Explain how AI suggestions are generated and keep the parent in control—allow easy edits and confirmation before publishing.
•
Close the feedback loop:
Let users easily share quick input so the AI can learn and improve its accuracy, relevance, and tone over time.
Streamlining
viewing to inquiry
Reduce confusion and hesitation when viewing a listing, helping parents quickly understand key details and move confidently to the next step.
•
Make key info easy to find:
Show essential details—condition, size, age range, and pickup location—clearly to help parents decide faster.
•
Provide a reference link:
Offer helpful context to support decision-making.
•
Show clear pickup location:
Use a nearby intersection or landmark with a small map view to indicate the approximate pickup area safely.
•
Clarify next-step actions:
Use clear copy and a friendly, pre-filled message to help users start an inquiry with minimal effort.
Improving Pickup Efficiency
Reduce the back-and-forth between givers and receivers with a built-in scheduling tool that simplifies arranging pickups while keeping communication personal.
Giver
•
Design an intuitive date and time picker:
Minimizing friction by letting givers quickly share available time slots in a few taps.
•
Promote communication efficiency:
Guide givers to provide pickup address and any special notes up-front to reduce repetitive messages.
Receiver
•
Enhance communication without losing the human touch:
The scheduling tool supports, not replaces, conversation—allowing users to confirm details or express gratitude in their own words.
•
Provide a clear pickup summary:
Surface confirmed time, location, and contact details so both sides have shared expectations.
Designing a User-Centered Dashboard
Create a central place where parents can easily track their exchanges, manage pickups, and discover new listings that match their interests.
•
Personalize the experience:
Use settings like location, children’s age, and interests—plus past activity—to tailor listing recommendations.
•
Simplify pickup management:
Give users a quick overview of upcoming pickups with easy actions to view, reschedule, or add to calendar.
Help parents easily find what they need with intuitive search and filters that adapt to their location, child’s age, and item preferences.
•
•
Building the prototype through vibe-coding came with constraints—some desired features couldn’t be implemented, and a few UI details required trade-offs. I focused on what mattered most: ensuring core user actions were intact to validate usability and user value.
User Testing & Feedback
I conducted five moderated user testing sessions to evaluate the impact of my design decisions. Through qualitative feedback and observation, I assessed ease of use, task completion, and friction points to understand usability, efficiency, perceived value, and overall satisfaction.
•
•
“
It was helpful because it gave me a starting point so I didn’t have to start from scratch.
“
It saves me a lot of time. I usually need to Google info manually.
•
•
“
I really like the time slot selection design and the summary at the end!
•
•
•
future opportunities
Beyond the current prototype, here are areas I’m interested in exploring further.
Adaptive Homepage
•
•
Connection Building
•
•
•
Contextual Listing Creation
•
•
Predictive Messaging
•
•
•
reflection & learning
•
Vibe-coding taught me to balance creativity with constraints.
I learned to define what truly needed to be prototyped and where smart workarounds could still tell the story effectively.
•
Iteration often led to unexpected insights.
While some ideas felt impossible to vibe-code after countless prompting rounds, other times the AI’s output surprised me with new angles I hadn’t first considered.
•
Clarity of intent matters more than precision of tools.
As technology keeps evolving, I’m reminded that strong design principles and clear goals ground the process—no matter the medium.
•
Staying open and adaptive fuels better design.
Embracing uncertainty and experimentation has become part of my growth mindset as a designer working at the intersection of UX and emerging technology.
