Skip to main content

Case study · SaaS / Social commerce

ChatcartOrder management for shops that sell on WhatsApp and Instagram

Chatcart is an order management web app built by DanixSoft for sellers who take orders over WhatsApp, Instagram and Facebook: it turns a pasted customer chat into an order, tracks each parcel from confirmed to delivered, and reconciles cash-on-delivery payouts from couriers.

A phone-first order book for social-commerce sellers in Pakistan. Paste a customer message and it becomes an order; move orders through packing and courier booking; record what the courier paid out against the delivered orders it covers.

What problem does Chatcart solve?

Sellers who sell through DMs keep orders in chat threads, notebooks and spreadsheets, so orders get missed, returning customers are not recognised, and nobody knows how much cash on delivery the courier still owes.

How does Chatcart work?

A rule-based parser pulls the name, Pakistani mobile number, city, address and items out of a pasted chat for the seller to confirm. Orders move through a fixed pipeline with bulk changes and undo, customers are recognised by phone number, and each courier payout is recorded against the exact delivered orders it settles.

What we built

  • Smart paste: customer chat to order, including WhatsApp chat exports
  • Order pipeline from new to delivered, plus returned and cancelled, with bulk moves and undo
  • Cash-on-delivery payout reconciliation per courier
  • Customer history by phone number, with a flag for frequent returns
  • Packing slips two to an A4 page and a CSV export for courier bulk booking
  • WhatsApp confirmation and dispatch messages opened in the seller's own WhatsApp

What is Chatcart built with?

Application
Next.js 16 (App Router, Server Actions), React 19, TypeScript, Tailwind CSS v4
Data
Supabase Postgres with row-level security and Auth
Quality
Vitest unit tests and a row-level security check script

Engineering challenges

  • Every table scoped to a store by row-level security, with column-level grants so totals and store ownership cannot be edited directly
  • Order creation and payout recording in single database transactions through security-definer functions
  • An append-only order event trail written by triggers
  • CSV export guarded against spreadsheet formula injection

Next step

Have a similar problem to solve?

A free 30-minute discovery call, then a written scope and a fixed-price quote — before you commit to anything.