Title: Cartify - The AI Powered Task-to-Cart Agent for Every Shopper

Inspiration

We've all been there.

You open your shopping app, just trying to cook biryani or prep for a house party. But then 23 ingredients, 10 YouTube videos, and 14 tabs later, you're frustrated, confused, and still forget the milk.

Shoppers don't want filters and categories. They want help. They want something that understands the task behind the purchase.

Ecommerce stores are struggling too, carts are abandoned, budgets blown, essentials forgotten. Shoppers waste time. Merchants lose revenue.

We weren't just building a feature. We were solving a human emotion, the gap between "I want to..." and "I bought it."

That's why we created Cartify , your shopping copilot.

What it does

Cartify transforms a user's task like "Make biryani under $500" into a complete, ready-to-checkout cart on any ecommerce site.

With one line of input, Cartify:

Understands what you need (ingredients, items, quantities)
Understands how you feel (lazy Sunday? busy Monday? festive mood?)
Stays within your budget
Saves your habits into reusable smart bundles
Lets you reorder anything in one tap

It works across any ecommerce store. It's fast, smart, and deeply personalized.

From forgetful shoppers to merchants struggling with low cart size, Cartify solves both.

How we built it

Frontend SDK: React + Tailwind CSS A lightweight SDK embeddable in any ecommerce store (Cartify's core experience)
Cartify Agent: Groq API + Llama 3.5 Converts user goals like "Snacks for $300" into structured JSON cart formats
Backend: Supabase (auth, DB, bundles) Handles user sessions, cart memory, API requests, and analytics events
Edge Functions: Performs smart budget logic, seasonal detection, mood analysis, cart generation triggers
Analytics Engine: Real-time dashboards in the Merchant Portal Tracks bundle popularity, average cart value, top-selling SKUs, etc.
Memory Layer (Cartify Memory): Stores smart bundles, reorder patterns, nudge triggers, seasonal restock logic

Challenges we ran into

API Rate Limits: The ecommerce API we used had a limited number of search requests. Once we exceeded it, we had to constantly re-authenticate with alternate accounts , slowing our build cycle.
Agent ↔ App Integration: The Cartify Agent generated cart data in JSON, but aligning that with the frontend SDK and product schema was tricky. Schema mismatches caused bugs until we stabilized the format.
Cart Memory System: Saving and retrieving bundles based on user, season, and context was more complex than expected. We had to carefully design Supabase tables and edge logic for privacy and speed.

Accomplishments we're proud of

Built a complete shopping agent - fully AI-powered, context-aware, and store-ready
Cart generation happens in under 2 seconds via Groq's blazing speed
Created reusable Smart Bundles that remember and re-trigger weekly
Developed a real-time dashboard with trend and conversion analytics
Embedded intent + emotion + budget understanding into shopping

What we learned

Users don't want to browse , they want outcomes. Task-first shopping is the future.
Connecting agents to real actions (cart, reorder, checkout) is 90% of the work
Memory is not a luxury , it's the bridge to loyalty