Inspiration

TideCarts was inspired by the challenges small business on remote islands face when trying to sell handmade goods globally. We wanted to help them sell their handmade products globally, to consumers from all over the world.

What it does

It is an e-commerce platform containing three connected engines including tidePrice, tideRoute, tideMatch. Changing the destination or shipping method updates the landed cost and pricing recommendation. TideCart explains each recommendation by showing expected demand, maker profit, landed margin, elasticity, confidence, and any constraint affecting the result. Customers can also add products to a working cart, adjust quantities, remove products, view route-aware totals, and reserve a shared shipment. The hackathon checkout stops before payment and collects no personal information.

How we built it

This was built with React 19, TypeScript, Tailwind CSS, shadcn-style UI components, Vinext, and Cloudflare Workers-compatible tooling. All three engines run locally in the browser with deterministic sample data, so judges can test the full experience without API keys or a separate Python server.

Challenges we ran into

One major challenge was connecting pricing and shipping meaningfully. Simply displaying shipping next to the recommended price was not enough—the route cost needed to participate directly in the pricing calculation

Accomplishments that we're proud of

Built a complete seller-to-customer flow instead of an isolated algorithm demonstration. Made shipping costs directly affect the dynamic pricing recommendation. Added seller protections through margin and price-change guardrails. Clearly disclosed infeasible pricing conditions instead of hiding them.

What we learned

We learned that an optimization result is only useful when the user understands why it was recommended. Showing one “optimal” number can feel like a black box, especially when a seller’s livelihood is involved. Explanations, constraints, and seller-controlled guardrails are essential parts of the product.

What's next for TideCarts

The next version of TideCart would: Connect to live carrier quotes and shipment schedules. Replace the deterministic pricing adapter with the upstream Fast API and trained machine-learning models. Learn elasticity from real seller demand and transaction history.

Built With

Share this project:

Updates

Submission history