Aeris
They approached us with a challenge, "How can we bring the magic of fragrance discovery to life online — powered by AI, yet crafted with emotion?"
Key Challenges®
//04Every innovation project brings its own set of complexities, especially when AI, user experience, and emerging technology intersect. In this project, our biggest challenge was to translate intelligent systems into human-centered interactions that felt natural, intuitive, and trustworthy.
No Sense of Smell Digitally
We had to design around the absence of smell — recreating its emotional impact through story, visuals, and interaction.
Fragrance Language is Abstract
Most users struggle to understand terms like "aldehydic" or "chypre" — the system needed to translate those into human feelings.
Avoiding Generic Quizzes
Many perfume finders feel robotic or superficial — we had to make the process feel artistic, intimate, and enjoyable.
Balancing AI with Brand Sensibility
Aeris is rooted in French luxury — the AI experience had to match that level of detail, storytelling, and refinement.
Design Approach®
//03We approached this project like designing an interactive scent ritual, led by AI but rooted in storytelling and emotion.
The Scent Muse — AI Discovery Guide
We created a conversational assistant named “The Muse” — a poetic, soft-spoken AI guide that asked users questions like, “What memory brings you peace?”
Olfactory Archetype Engine
Each perfume recommendation was presented with immersive visuals, poetic microcopy, and even ambient soundscapes for emotional resonance.
Sample Kit Personalization Flow
After discovery, users could build a bespoke sample kit with AI-selected fragrances — accompanied by printed scent stories, rituals, and a guide for layering.
Final Outcome
//04The result was a seamless, intelligent experience that merged advanced AI capabilities with thoughtful, human-centered design.
Users described the discovery journey as “memorable” or “emotionally resonant”
increase in sample kit orders through AI-recommended paths
reduction in return rates from first-time buyers who used the Muse
Time on page tripled compared to standard product flows