AUROMA
FIND THE AROMA TO YOUR AURA
Objectives

Personalisation
Takes your unique preferences and needs into account.

Immersive
An experiential and creative journey tailored to you.

Bridging Understanding
Helps align your perception of fragrances to the fragrances’ actual smell.

Shareability
Designed to be shared, fostering connection through fragrance.
User Experience






Core Features
Profiling
Embark on a journey. As you navigate engaging scenarios, you will unpack and map your personal fragrance preferences to uncover your identity.


Exploration Map
With a selection of over 200 fragrances across 7 brands, get matched with your signature fragrance – most tailored to your olfactory preferences – with our vector similarity driven recommendation engine.
Additionally, explore mood-based recommendations curated for you that reflect and evoke the core emotions that define your unique essence.
Product Pages
Deep dive into the fragrance of your choice with all the information you need to make an informed decision.
Grasp a fragrance’s true scent with an overview of vital information and visualisations of notes. Head to the Reviews tab to view text reviews condensed into concise topical summaries powered by LLMs, and curated social media video reviews tailored to your archetype personality.
Intrigued? Check out the fragrance on the official website or book an in-store appointment with convenience.
Results and Validation
Users like their recommendations

4.34 out of 5
Prefer AUROMA to other fragrance finders
Helped 82% of users understand fragrance terminology better




System Architecture
AUROMA is deployed on Google Cloud with React as the frontend and FastAPI as the backend, supported by TaskIQ and Redis for asynchronous workflows. AUROMA’s AI components include a pre-trained decision model and Google VertexAI.

Acknowledgements
We would like to sincerely thank our Singapore University of Design and Technology (SUTD) Capstone mentor, Professor Qun Song, and our communications advisor, Dr. Susan Wong, and our L’Oréal industry mentors, Nicol Ong and Jude Low, for their guidance and support throughout the entirety of this Capstone project.
We would also like to extend our gratitude to the project supporters, Ashley Zhao, Zoey Lai, Harnyanuvat Ann Yolyupa, Sebastian Au, Bianca Taiwo & Sophie Frakes, for their contributions and consistent support towards our project.