Inspiration
My inspiration came from a shared, frustrating experience: the overwhelming and often intimidating process of choosing a new perfume. Walking into a store, you're faced with a "wall of scents", confusing terminology, and pressure to make an expensive choice quickly. I realised this was a perfect problem for AI to solve. We envisioned a personal, expert consultant that could sit with you, understand your unique personality and needs, and translate that into the perfect scent. The idea was to bring expertise and personalisation to everyone, demystifying the beautiful world of fragrance.
What it does
FragranceFolio is a smart, AI-powered web application that acts as a personal fragrance consultant. It simplifies scent discovery through a simple, conversational experience:
Users answer a few intuitive questions about their preferences, the occasion, desired mood, and scent notes they enjoy.
Our AI, powered by a single, sophisticated prompt, analyses their answers in real-time against a database of over 70,000 fragrances.
It instantly delivers a top recommendation with a detailed, personalised explanation of why it's a perfect match, along with a curated list of the top 10 alternatives for further exploration.
How I built it
Data: I used a comprehensive CSV dataset (fra_perfumes.csv) containing over 70,000 fragrances and their detailed attributes. Frontend: I designed a minimalist, mobile-first user interface using a high-performance web framework like Next.js to ensure the application is incredibly fast and responsive.
Challenges we ran into
Crafting a single prompt that could reliably handle the nuanced logic of fragrance recommendation was our biggest challenge. It took dozens of iterations to perfect its ability to interpret user intent, weigh different factors correctly, and generate consistently high-quality, coherent responses.
Accomplishments that we're proud of
I am incredibly proud of successfully building a fully functional and complex application powered by a single, elegant prompt. This demonstrates a new, incredibly rapid, and flexible paradigm for AI app development.
What's next for FragranceFolio
User Profiles & Scent History: Allow users to create accounts, save their favourite fragrances, and view their recommendation history.
Community Integration: Introduce features for user reviews, ratings, and community-curated lists to add a social layer to scent discovery.
E-commerce Partnerships: Integrate with retailer APIs to provide seamless "check price" or "buy now" options for recommended products.
Expanding the AI's Senses: Enhance the questionnaire with more abstract, visual, or mood-based questions to create even more nuanced and accurate scent profiles.
Native Mobile Apps: Develop dedicated iOS and Android applications for an even richer mobile experience.

Log in or sign up for Devpost to join the conversation.