Synclo

Synclo uses Gemini 3 to digitize your closet with one photo, suggesting outfits based on weather, trends, and schedule. Generating a realistic Virtual Try-On image of you wearing the combination.

Comment

Inspiration

Synclo was born from observing my wife's frustration every morning at not knowing what to wear to her social or work events. And my own lack of fashion sense to suggest anything to her.

What it does

Ever felt like you have a closet full of clothes but nothing to wear? Synclo is your AI-powered personal style agent.

Synclo is a solution that leverages Gemini 3’s reasoning to automate personal image management through the following workflow:

Smart Digitization: The user uploads a photo of a garment, Gemini 3 executes image segmentation to remove the background and mount the item on a hanger, creating a visually flawless and unified catalog.

Automatic Tagging: Simultaneously, Gemini 3 performs a technical classification, identifying textures, color, and styles to tag each item in the database.

Contextual Reasoning: When requesting an outfit, Gemini 3 analyzes the entire inventory and synthesizes external variables—such as local weather, current trends, and specific events—to propose the optimal garment combination.

Hyper-realistic Visualization: In its most powerfull phase, Gemini 3 generates a hyper-realistic image of the user wearing the suggested combination, allowing them to virtually validate their appearance before physically getting dressed.

How we built it

Our development methodology began with a "Vibe coding" phase in Google AI Studio before migrating the entire environment to Antigravity.

  1. AI Engine: We integrated Gemini 3, leveraging its vision for image segmentation, automated technical tagging, and the reasoning required for generating Virtual Try-On (VTO) images.
  2. Backend: We utilized Firebase as our core infrastructure. We implemented Firestore for real-time data management and Firebase Auth to ensure a secure and private user environment.
  3. Frontend: The system features a hybrid architecture: a web application developed with React and a mobile version (currently in progress) using React Native, ensuring a consistent user experience across any device.
  4. Execution Logic: We configured Gemini 3's thinking_level parameter to dynamically process and synthesize external variables, such as weather forecasts and the user's schedule, into contextual style recommendations.

Challenges we faced

One of the main characteristics of VTO image generation is that it consumes a lot of AI resources, increasing the app's cost. That's why we also implemented collage-type images for preliminary testing before moving on to a final VTO.

What we learned

Despite not having extensive programming experience, Antigravity's enormous reasoning ability allowed us to develop a clean and fully functional application architecture in record time.

What's Next for Synclo

Our short-term vision is to bring the Synclo experience to all mobile devices, completing the migration to React Native to publish the application on the iOS and Android stores.

Our strategy is divided into two key phases:

  • Mass Acquisition: We will launch the application with a highly competitive or even free pricing model. The goal is to eliminate entry barriers and allow thousands of users to digitize their wardrobes, validating our AI technology.
  • Commercial Partnerships: Once a solid user base is established, we plan to forge paid collaborations with major clothing retailers. These brands will be able to digitize their stock within Synclo, enabling the app to suggest outfits featuring their garments and converting these recommendations into real purchases for each collaborating brand.
Share this project:

Updates