-
-
Page cover
Inspiration
We are an interdisciplinary team, and our project’s interface is inspired by the minimalist aesthetic that has defined Inditex brands in recent years. Following this philosophy, we designed FitTEX to be as intuitive and user-friendly as possible. Our goal was to leverage artificial intelligence to simplify the user experience, making fashion discovery effortless and efficient.
What it does
FitTEX offers an advanced search experience with two approaches:
- Image-based search: Upload a photo, and the system will find similar garments from the Inditex catalog.
- Text-based search: Describe what you're looking for in natural language, and our AI model will interpret your request to suggest the best matches.
How we built it
From a technical perspective, we utilized a wide range of technologies to apply our academic knowledge and professional experience. One of the biggest challenges was fine-tuning the LLM, as we had no prior experience in this area. This required extensive research, experimentation, and learning to achieve optimal results.
What we learned
This project allowed us to deepen our expertise in two key areas:
- Deploying applications with multiple technologies, ensuring smooth integration between backend, frontend, and AI components.
- Fine-tuning large language models (LLMs), an entirely new area for us, where we gained hands-on experience in data preprocessing, model optimization, and adapting AI to specific use cases.
What's next for FitTEX
The next steps for FitTEX focus on two main areas:
- AI Model Optimization: While fine-tuning has significantly improved performance, it remains a resource-intensive process. Finding more efficient ways to enhance model accuracy will be a key priority.
- Frontend Evolution: We aim to continue refining the user interface, maintaining our minimalist philosophy while integrating new functionalities that further enhance usability and accessibility.
Log in or sign up for Devpost to join the conversation.