Inspiration
We wanted to build a personal AI stylist that truly understands body type, skin tone, and occasion. Everyone struggles to choose the perfect outfit for different events.
What it does
Our project generates personalized outfit recommendations with AI-generated images. Users input their body type, occasion, and preferences, and our agentic system creates stylish looks tailored for them.
How we built it
We built it using Google ADK (Agent Development Kit), Vertex AI, Imagen 3 for image generation, and Python. We designed a multi-agent workflow where agents collaborate to analyze style and generate visuals.
Challenges we ran into
Integrating multi-agent logic, fine-tuning Imagen 3 prompts for realistic outfits, and ensuring recommendations were truly personalized were major challenges.
Accomplishments that we're proud of
We successfully built a working agentic workflow with real-time image generation that provides practical and stylish outfit suggestions.
What we learned
We learned multi-agent system design using Google ADK, advanced prompt engineering, and how to integrate LLMs with image generation models.
What's next
We plan to add virtual try-on, wardrobe management, and shopping platform integrations to make it a complete styling assistant.
Built With
- google-adk
- google-cloud
- imagen-3
- python
Log in or sign up for Devpost to join the conversation.