Inspiration

Are you ever in a place but unsure of what you need? ARC'TERXY - OutfitConnect is here to enhance your options with vibrant choices and a seamless experience for all visitors, whether you are here to share your fun stories with our community or to look for your favorite item.

Our cutting-edge omnichannel app lets you effortlessly snap and flaunt your adventure-ready outfits, with instant access to every item for hassle-free shopping. Immerse yourself in a vibrant community of like-minded adventurers, where you can share tips, connect with fellow explorers, and ignite your passion for outdoor fitness, enhancing your overall wellness journey. Join us and elevate your outdoor experience to new heights!

What it does

  • Seamless Discovery: Capture any outdoor gear in a photo, and Arc'teryx's AI instantly identifies it, guiding you to similar products.

  • Share & Inspire: Post your outdoor adventures and let Arc'teryx highlight your gear, connecting you with a community eager to follow in your footsteps.

  • AI-Powered Match: No more guessing games; Arc'teryx provides direct links to gear matched by AI, making your next purchase a tap away.

How we built it

  • OpenAI: OpenAI API was used to do the Image Analysing and Product Matching
  • Auth.js: A secure authentication solution that guarantees user data privacy and security, ensuring your personal information remains confidential.
  • Next.js: A React framework that enables functionality such as server-side rendering and generating static websites for React-based web applications.
  • Tailwind CSS: A utility-first CSS framework for rapidly building custom designs, ensuring a sleek and intuitive user interface.
  • MongoDB: A NoSQL database that offers scalability and flexibility with querying and indexing, perfect for handling dynamic journaling content.

Challenges we ran into

  1. Unpredictable AI Outputs: The core challenge lies in the inherent unpredictability of large language model outputs, which might not always align with the desired format or specificity needed for accurate gear matching.

  2. Limited Product Data: A significant hurdle is the limited availability of comprehensive product data, which is crucial for AI to make accurate and relevant gear matches.

Accomplishments that we're proud of

Seamless User Experience: Users can easily post their activity as a blog, including sharing images of their outfit of the day (#ootd) and accompanying feelings. This creates a platform that is engaging and encourages user interaction.

Integration with OpenAI API: Implementing the OpenAI API enables the analysis of outfit images. By utilizing advanced image processing capabilities, users can receive detailed insights into their outfits, such as identifying clothing items, colors, patterns, and styles.

Real-time Image Analysis: The system provides real-time analysis of outfit images, ensuring quick and accurate results. This enhances user satisfaction by delivering prompt feedback on their fashion choices.

Personalized Recommendations: Leveraging the fake database image table, the system matches user input with relevant items and provides URLs for further exploration. This personalization enhances the user experience by offering tailored recommendations based on individual preferences and style.

Secure Authentication: Implementing robust authentication mechanisms ensures that user data remains secure and protected. By prioritizing security, the system instills trust and confidence among users, encouraging continued engagement with the platform.

What we learned

  1. Learn the concept of OAuth, and how to implement it into the project
  2. Use LangChain to do prompt engineering
  3. Using Next.js Server Actions to improve performance

What's next for ARC'TERXY - OutfitConnect

  • Expand AI Training Set: incorporating a broader spectrum of images and user interactions to improve both recognition and matching capabilities

  • Enhance Personalization: Develop algorithms that adapt to individual user preferences and behaviors, offering more tailored gear recommendations and a personalized user experience.

Built With

Share this project:

Updates