Try on AI - Revolutionizing Online Shopping

See it before you buy it — AI-powered Try-On & Try-In that brings products to life, from fashion to home décor.

Comment

🔹 Inspiration

Online shopping has a $800B+ annual returns problem.
Why? Because customers can’t truly see how a product will look on them or in their homes.

We asked ourselves:
“What if AI could close the imagination gap — letting you **try on clothes* or place décor in your room before hitting buy?”*
That question sparked this project.

🔹 What We Built

Our solution is an AI-powered Try-On & Try-In system:

  • Upload a selfie → preview sunglasses, shirts, or watches on yourself.
  • Upload a room photo → visualize candle holders, jars, or furniture inside your space.
  • All generated seamlessly & photorealistically using Google Gemini AI.

🔹 How We Built It

  1. Base Platform: Deployed Google’s Online Boutique microservices demo on GKE Autopilot.
  2. Custom Try-On Service: Flask backend + Gemini API to combine user images with product images.
  3. Integration: Extended Online Boutique’s frontend templates with a Try On button and upload form.
  4. Infra & DevOps:
    • GCP Cloud Build + Artifact Registry for CI/CD.
    • Kubernetes manifests for scalable deployment.
    • LoadBalancer for external access.

🔹 What We Learned

  • How to orchestrate microservices on GKE Autopilot.
  • Best practices for integrating AI into cloud-native apps.
  • Prompt design for photorealistic image generation.
  • Handling cross-service communication and secure API integration.

🔹 Challenges We Faced

  • Image alignment & realism: Ensuring products looked natural (e.g., glasses on a face vs. candle holder on a table).
  • Kubernetes integration: Debugging pod networking + service exposure.
  • API limits & latency: Optimizing requests so the experience feels instant.
  • Frontend integration: Embedding a new Flask service seamlessly into an existing Go/HTML microservice frontend.

✨ In short:
This project is more than a demo — it’s a proof-of-concept for the future of e-commerce, where AI + Cloud eliminate uncertainty and create confidence-driven shopping.

Share this project:

Updates