Inditex Visual Search

Find fashion pieces similar to the ones you love with Inditex’s powerful visual search technology. Just upload a photo or snap one with your camera—and discover matching styles in seconds.

Comment

🧠 Inspiration

We were really intrigued by the model developed by the Inditex Tech team, which allows users to find similar clothing items from their brands just by uploading an image. What caught our attention the most was the opportunity to experiment with image preprocessing to improve the performance and accuracy of the AI behind it.

--

🛍️ What It Does

Our application enables users to:

  • 📷 Upload an existing photo or take one using their device’s camera
  • 🧠 Search for visually similar clothing items using Inditex’s visual search technology
  • 🎯 Interact with a UI designed to closely match the simplicity and elegance of Inditex’s platforms

The goal: make the experience intuitive, fast, and enjoyable for any user.

🛠 How We Built It

Our development journey involved several stages:

  • 🔄 Started with React + FastAPI, but later migrated to a full Next.js setup for simplicity and better integration
  • 🚀 Deployed the app on Vercel for seamless hosting
  • 🧩 Used Server-Side Rendering (SSR) to protect API credentials and handle secure calls
  • 🖼️ Experimented with multiple web scraping techniques to extract product images
  • 🔐 Implemented token auto-refreshing to ensure API access remained stable in production

We also designed a clean, central interface to guide users through the entire process effortlessly.

🧗‍♀️ Challenges We Ran Into

This wasn’t as simple as it first seemed. Some key challenges:

  • 📉 Communication hurdles slowed alignment across the team
  • 🕷️ Vercel doesn’t support tools like Puppeteer, so scraping product images became difficult
  • 🛡️ Zara’s anti-bot security blocked us in production (though we bypassed it successfully in local environments!)
  • 🧪 API calls failed due to hidden character issues—solved thanks to help from Alexandre Ortega (Inditex Tech)

Despite these setbacks, we kept iterating and found workarounds to push forward.

🏆 Accomplishments We're Proud Of

  • ✅ Built a functional MVP under time constraints
  • ✅ Created real value with a polished user interface
  • ✅ Solved technical obstacles through teamwork and persistence
  • ✅ Delivered a product that matches the challenge’s goal

This wasn’t easy, but we’re proud of what we accomplished!

📚 What We Learned

  • 🔄 Refreshed our Next.js and frontend skills
  • 🕸️ Learned advanced web scraping under production constraints
  • 🌐 Gained experience deploying full-stack apps on Vercel
  • 🎯 Improved our ability to scope ideas and prioritize features

We started big, but adapted quickly and focused on what really mattered.

🔮 What’s Next for Inditex Visual Search

We have tons of ideas we’d love to implement next, such as:

  • 🧪 Image preprocessing to boost AI accuracy
  • 🧍 Virtual try-on or 3D garment previews
  • 🔍 Detecting and selecting individual clothing items within a photo
  • 🏪 Checking product availability in nearby stores (when API access allows)
  • 👤 Adding user accounts for personalized recommendations based on previous searches

Built With

Share this project:

Updates