Recently, the HackXelerator program concluded with a showcase of next-generation, AI-driven innovation. Among the standout projects was StyleGenie, the winner in the Fashion category; a personal AI stylist that merges fashion-forward thinking with technical excellence to solve real-world industry challenges.

A more innovative way to style

StyleGenie isn’t just another recommendation engine – it’s an AI-native platform that delivers complete outfit suggestions personalized to the user’s style, budget, body type, and occasion. Built during a 20-day creative sprint, the project combines cutting-edge AI tools with a strong product vision and sustainable values.

Unlike traditional e-commerce platforms that recommend single items, StyleGenie curates ccomplete looks. Users can visualize and virtually try on outfits, tweaking them to suit different occasions or moods. For brands, it’s a step toward hyper-personalized customer experiences, fewer returns, and smarter inventory decisions.

Meet the team

• Juliana Varela: Creative and fashion lead
• Garret Luo: Product lead
• Leonardo Jacobi: Backend, DevOps, and AI lead
• Nada Kartouch: Frontend engineer

The team formed organically through the HackXelerator Discord server, united by a shared passion for AI and innovation in fashion.

How StyleGenie came to life

Garret first proposed building an AI-powered fashion recommendation system. Juliana expanded on the idea by integrating virtual try-on, and the team quickly coalesced around a unified vision: an intuitive AI stylist that bridges creativity and utility.

They began by defining the product strategy, brand identity, and technical roadmap. Over the next three weeks, they layered in powerful AI components like:

• Local Ollama API and LightLLM for generative capabilities
• CrewAI for agent orchestration
• LangGraph for AI workflows
• Google ADK and A2A for enhanced automation
• FastAPI, Python, React, and Websockets for backend and frontend development
• MCP and MCP-Bridge for connecting system components

The result was a scalable, AI-native architecture designed for experimentation and real-world deployment.

Powered by Vultr

To support their infrastructure, the StyleGenie team turned to Vultr. This gave them the flexibility to:

• Deploy and test AI models in a real-world environment\
• Seamlessly connect local APIs with cloud-based components
• Scale their backend as needed during the development sprint

One of their key breakthroughs came from gaining access to an AMD GPU, enabling them to run models like Ollama and LightLLM locally. This reduced dependency on third-party APIs and laid the groundwork for future scalability.

The HackXelerator effect

With support from Vultr, KXSB, and Pinecone, the team benefited from infrastructure, mentorship, and deep technical guidance. Community connections and events in London, Paris, and Berlin supercharged their momentum and opened the door to future collaboration.

“The support and structure of HackXelerator allowed us to move fast, test boldly, and bring our vision to life,” said the team. “The constant innovation and encouragement made the experience deeply rewarding.”

What’s next for StyleGenie?

The team is just getting started. Their roadmap includes:

• Formalizing the infrastructure for anonymous AI agents
• Building analytics and marketing tools
• Launching a beta site for accurate user testing\
• Partnering with fashion brands
• Exploring funding opportunities to scale

StyleGenie is more than a hackathon project: it’s a bold step toward reimagining how people shop for fashion in a digital world.