Mira — rethinking fashion discovery through conversation
Exploring how conversational AI can make fashion discovery more intuitive, turning intent into relevant product recommendations through natural language.
problem
Fashion ecommerce faces a double challenge: attracting qualified traffic is becoming harder, while turning visits into purchases still depends on helping users quickly find something they actually want. When discovery relies on endless browsing, categories and filters, intent can easily get lost before conversion happens.
solution
A conversational shopping experience that translates natural language into relevant product recommendations, helping users move faster from intent to discovery and reducing friction before purchase.
Context
Fashion ecommerce is increasingly competitive: acquiring traffic is harder, attention is shorter, and conversion depends on how quickly users find something relevant. At the same time, shopping intent is often contextual — an occasion, a mood, a budget or a desired look — while most ecommerce experiences still ask users to translate that intent into categories and filters.
Hypotheses
H1 — Natural language can reduce discovery friction
If users can describe what they need in their own words, they may reach relevant products faster than through traditional categories and filters.
H2 — Contextual recommendations can increase relevance
If the experience understands occasion, budget, style and intent, recommendations may feel more useful and reduce unnecessary browsing.
H3 — Better discovery can improve purchase intent
If users find relevant products earlier in the journey, they may be more likely to explore product details, consider a purchase and continue towards conversion.
Design challenge
How might we help users move from a vague shopping intention to relevant fashion products faster, without forcing them to translate what they want into categories, filters and search terms?
year
2026
timeframe
2 weeks
tools
Figma · Framer · ElevenLabs · Supabase · ChatGPT
category
Personal Project