AETHER~ Fashion E-Commerce Platform
Budget: ₹1,500 – ₹12,500 INR
I’m building AETHER, a MERN + Tailwind digital wardrobe that merges vintage and pop-culture pieces with real-time AI styling. The storefront is already scaffolded; what I need now is the brains behind it: an AI stylist that can look at a shopper’s body data and instantly suggest outfits from our catalogue.
Scope
• Focus of the model: body type analysis at an intermediate level—pear, apple, rectangle, inverted triangle, and similar shapes.
• Workflow: a user enters height, weight, and a few key measurements; the service classifies their shape, queries MongoDB for matching products, then serves up complete looks ranked by relevance.
• Stack: React on the client, Express/Node API, MongoDB Atlas, Tailwind CSS for styling. You’re free to slot in OpenAI, TensorFlow, or another framework as long as the final service is container-ready and callable from Node.
• UX touchpoints: a React component that renders recommendations inline, plus an admin view so I can tweak shape-to-style mappings without redeploying.
Deliverables
1. AI classification service (Docker image or serverless function)
2. Secure API endpoints and Mongo queries that feed the stylist real-time inventory data
3. Front-end components with Tailwind styling and loading states
4. Setup docs and a short test script that proves shape detection and product matching work end-to-end
Acceptance Criteria
– Accurate assignment to the intermediate body-shape list above
– Outfit recommendations returned in <2 seconds for a catalogue of ~5 k items
– All code lint-clean and pushed to my GitHub repo with clear README
If you’re comfortable marrying fashion insight with solid MERN engineering, let’s get AETHER’s stylist dressing users in style.
Scope
• Focus of the model: body type analysis at an intermediate level—pear, apple, rectangle, inverted triangle, and similar shapes.
• Workflow: a user enters height, weight, and a few key measurements; the service classifies their shape, queries MongoDB for matching products, then serves up complete looks ranked by relevance.
• Stack: React on the client, Express/Node API, MongoDB Atlas, Tailwind CSS for styling. You’re free to slot in OpenAI, TensorFlow, or another framework as long as the final service is container-ready and callable from Node.
• UX touchpoints: a React component that renders recommendations inline, plus an admin view so I can tweak shape-to-style mappings without redeploying.
Deliverables
1. AI classification service (Docker image or serverless function)
2. Secure API endpoints and Mongo queries that feed the stylist real-time inventory data
3. Front-end components with Tailwind styling and loading states
4. Setup docs and a short test script that proves shape detection and product matching work end-to-end
Acceptance Criteria
– Accurate assignment to the intermediate body-shape list above
– Outfit recommendations returned in <2 seconds for a catalogue of ~5 k items
– All code lint-clean and pushed to my GitHub repo with clear README
If you’re comfortable marrying fashion insight with solid MERN engineering, let’s get AETHER’s stylist dressing users in style.