Qaemah: No-Code Store Builder with WhatsApp Ordering

Startup

A no-code platform that turns any small business's products into a polished online store on its own web address, with one-tap WhatsApp checkout and a built-in, catalog-aware AI shopping assistant that speaks Arabic and English.

NextJSReactJSTypeScriptTailwind CSSPostgreSQLAI Integration

Overview

Qaemah (Arabic for "list" or "menu") lets any small business turn its products into a polished online store in minutes: no developer, no heavy e-commerce platform. The owner adds their products and Qaemah instantly gives them a public, mobile-friendly storefront on their own web address; when a customer wants to buy, they skip the complicated checkout and place the order straight through WhatsApp, the app these businesses and their customers already use every day. It's built Arabic-first (full RTL, Arabic + English throughout) for markets where millions of shops have a WhatsApp number and a phone full of product photos, but no proper catalog.

The Challenge

Those shops lose sales to messy back-and-forth messages and screenshots. The goal was to give them a real storefront without changing how they (or their customers) already work, and to make it genuinely no-code, multi-tenant, bilingual, and smart.

  • Genuinely no-code for a non-technical owner, yet every store still feels custom-branded and lives on its own domain.
  • Multi-tenant: one platform quietly running many independent stores side by side, each fully isolated.
  • Flexible enough for wildly different businesses: a grocer selling by weight, a café with add-ons, a fashion shop with sizes.
  • First-class Arabic RTL, plus a shopping experience smart enough to actually help customers find things.

How I Solved It

Qaemah is a modern Next.js + React + TypeScript web app (Tailwind + accessible Radix UI), talking to a separate backend API over PostgreSQL and installable as a PWA. Multi-tenancy gives each business its own store, team, branding and address (mystore.qaemah.com or a custom domain). The product model is deliberately flexible: sell by fixed price, by weight ($5/kg with custom quantities), or by size/variant (Small $6, Large $12); add customer-selectable options; and define your own product attributes with custom field types: the forms and tables adapt automatically. Bulk Excel/CSV import/export plus barcode and on-device OCR helpers get a shop with hundreds of items set up fast.

The standout is the AI Shopping Assistant built into every store: a chatbot that streams replies as it types, understands the store's actual catalog through semantic (vector) search powered by pgvector, works across Arabic and English (an Arabic question can surface English-labeled products), remembers the conversation, and shows interactive product cards you can add to the cart right from the chat. Four roles (Super Admin, Owner, Manager, Staff) gate access across every screen, and the app ships with unit and end-to-end tests (Vitest + Playwright).

Key Features

  • No-code storefront builder with full custom branding (logo, colors, fonts, card layouts, banners) and per-store domains.
  • One-tap WhatsApp checkout: orders arrive as ready-to-read messages; no payment gateway to set up.
  • Flexible pricing (fixed / by weight / by size), customer-selectable product options, and custom field types.
  • Bulk Excel/CSV import/export, barcode lookup, and on-device OCR for fast data entry.
  • Mobile-first storefront with search, category filtering, image galleries, a per-store cart, and multi-currency display.
  • Built-in AI shopping assistant: bilingual, catalog-aware, conversational, with add-to-cart product cards.
  • Multi-tenant SaaS with role-based access, subscriptions (30-day trial, Starter $9, Pro $14, Custom), and an installable PWA.

Results & Impact

Qaemah collapses "I want to sell online" from a multi-week, developer-dependent project into a self-serve setup an owner finishes alone, and by leaning on WhatsApp instead of a checkout, it removes the single biggest barrier to going live for small regional businesses. Its scope is real: one platform running many independent stores, a full suite of management modules across the dashboard plus a complete public storefront, role-based permission levels, two languages with full RTL throughout, multiple pricing models, custom field types, and a production-grade, bilingual, catalog-aware AI assistant built into every single store.

More Projects