Experimental Project · 2025
Redefining E-Commerce - AI shopping
Experiment Conversion Rate
- Role
- Product Owner, UX Designer & Prompt Engineer
- Duration
- 2 Weeks
- Year
- 2025
- Client
- Experimental Project
Overview
Project overview
Today’s online stores present products. But what if they could guide customers the way a personal stylist would?
This concept explores a smarter e-commerce flow designed to support different shopping behaviors.
- The problem
- Traditional e-commerce platforms offer browsing, filtering, and checkout; But the experience often feels transactional and impersonal.
- The decision
- The final concept transforms a static shopping site into a dynamic, personalized experience: Dual shopping modes that respect user preference and intent AI-powered assistant offering curated…
The challenge
Understanding the problem
Traditional e-commerce platforms offer browsing, filtering, and checkout; But the experience often feels transactional and impersonal. Users struggle with choice overload, poor product relevance, and time-consuming search flows.
Key problems surfaced:
Key problems surfaced:
- Shoppers spend too much time narrowing down options
- Product discovery relies heavily on manual filtering
- Search results lack context and personalization
- Users often abandon checkout due to friction and unnecessary steps
The challenge was clear: create an experience that feels more guided, intuitive, and human without adding complexity.
The approach
How I approached it
- 01
Started by exploring how AI could improve the shopping experience. I tested ideas by creating and refining prompts using Gemini to see how an assistant could give smarter recommendations.
- 02
Prototyped two parallel flows using Lovable: Shop on your own for traditional exploration & Assist me for AI-powered guidance.
- 03
Explored multimodal interactions
questions, natural language input, and image upload to support different mental models.
- 04
Identified Armani Exchange as the key design inspiration, adopting cues from its clean layouts, balanced visual hierarchy, and modern product presentation.
The solution
What I designed
The final concept transforms a static shopping site into a dynamic, personalized experience:
- Dual shopping modes that respect user preference and intent
- AI-powered assistant offering curated recommendations, contextual search, and image-based styling
- Smart product discovery, reducing the steps required to find relevant items
- Context-aware checkout, automatically detecting location and simplifying required inputs
- A workflow that feels closer to interacting with a human stylist — but optimized through AI
- Automatically detects the user’s location to avoid long form fills during checkout.
The result is an elevated e-commerce model: more intuitive, more efficient, and meaningfully more personal for every shopper.
Visual designs
Design highlights
01 / 02The Vibe Coding Process