Inspiration
Growing up, I was always fascinated by the intersection of fashion and technology. Two of my favorite shows were Totally Spies, where the characters had gadgets cleverly embedded into their clothing, and Kim Possible, with Kim’s futuristic tech bodysuit and spy gear. But what really sparked the idea for StyleBuddy was Clueless, specifically, Cher Horowitz’s iconic digital closet that effortlessly organized and styled her outfits. I remember thinking, “What if something like that existed today, something smart, but on a more personal, everyday level?” That question planted the seed for Outfit Genius: a tool to bring that futuristic, effortless styling experience to life for real people.
What it does
It helps you organize your wardrobe digitally, suggests outfits based on what you own, the weather, and the occasion or vibe you're going for, and lets you save favorite looks for future reference. Using your phone’s camera, it can function like a virtual mirror, overlaying outfit ideas on your reflection so you can visualize combinations before trying them on. Outfit Genius takes the stress out of styling and helps you feel confident in your choices, every day.
How we built it
I built OutfitGenius using Firebase for the app’s backend and Gemini AI for its intelligence. Firebase powers secure user authentication, real-time wardrobe data storage with Firestore, image hosting with Cloud Storage, and serverless Cloud Functions for generating outfit suggestions and sending notifications. Gemini AI handles vision tasks like detecting clothing types and colors from uploaded images, and uses its language capabilities to provide personalized outfit recommendations based on the weather, occasion, and mood. Together, these tools create a seamless, scalable app that acts like a personal stylist in your pocket.
Challenges we ran into
One of the biggest challenges I faced while building Outfit Genius was designing the vision system to accurately generate and display outfit cards that combined clothing pieces in a way that felt cohesive and intentional. While Firebase provided a solid backend for storing wardrobe data and serving images, it wasn’t designed for advanced vision processing or contextual outfit matching out of the box. It was challenging to balance automation with aesthetic logic to make sure that the combinations made sense stylistically, looked good visually on the card, and aligned with the user’s input on mood or occasion. This pushed me to explore more refined AI pipelines and custom logic for pairing items meaningfully.
Accomplishments that we're proud of
I used Gemini to refine and enhance the outfit generation process and overcome this challenge. Instead of relying solely on Firebase’s data retrieval, I used Gemini’s advanced language and vision models to create smarter prompts that could evaluate the wardrobe pieces contextually. I refined prompts within my Firebase functions to dynamically generate aesthetic outfit card layouts that visually balanced the pieces and felt intentional. This integration allowed Outfit Genius to present outfits that not only made sense technically but also looked polished and on-brand for the user’s chosen mood.
What we learned
Through building Outfit Genius, I learned how to effectively use AI to bring my product ideas to life. I gained valuable experience in integrating AI models like Gemini to solve complex design and logic challenges, and in using tools like Firebase to build scalable app infrastructure. It was exciting to learn how to “vibe code” to find creative, flexible ways to experiment, prototype, and iterate faster. This process taught me how to move from idea to implementation more efficiently, and fueled my passion for building innovative, user-centered solutions.
What's next for Outfit Genius
I hope to integrate a learning model to personalize outfit suggestions based on user-approved choices and adapt to location style preferences. Also, I want to create a feature for generating packing lists for trips, considering the destination's climate and cultural clothing expectations.
Log in or sign up for Devpost to join the conversation.