SORT IT OUT — Pan-India O2O Fashion Discovery & Multi-Tenant Commerce Platform

"Sort it, find it, shop it near you."

SORT IT OUT is a multi-tenant Online-to-Offline (O2O) platform that lets every independent boutique, regional chain, and ethnic/Western retailer in India run a storefront, while shoppers discover apparel near them by PIN code, filter ("sort") by what they want, and either buy online or walk into the physical store already knowing what's in stock.

What's actually running

apps/web is a real, working Next.js app — not just docs. Run it with:

cd apps/web && npm install && npm run dev

Consumer site (/)

• Hyperlocal discovery by PIN code, radius (1–25 km), and Men/Women/Kids category + subcategory filters (hover dropdown on desktop, tap accordion on mobile).
• Shop in Sort — cross-store product search (price, size, sort-by) that finds matching apparel across every nearby store at once. Save a filter combo as a Smart Sort (/sorts) and re-apply it later.
• Per-store storefronts (/store/[slug]) sharing one design system — header, fonts, spacing, and card style are uniform site-wide; each store's theme only tints its own hero gradient and CTA/accent color, so stores stay distinguishable without looking like different products.
• This platform doesn't sell online — it's O2O discovery. Wishlist items and picks go into Your Sort (/cart), then Reserve for Pickup (/reserve) holds them at that store for a chosen window (30 min–4 hr) so a shopper can walk in and try/buy in person; the seller is notified and sees it live in their Reservations tab. Track holds in My Reservations (/reservations), including a live pickup countdown.
• Live Sale — a seller can flag their store as running a flash sale right now (headline, discount, duration); shoppers see a live badge on the storefront, /shops, and homepage "Stores near you".
• Shopper login (/account) — mocked OTP sign-in (any 10-digit number, demo OTP is always shown on screen); once signed in, the header greets you by name and the reservation form/reviews prefill from the session.
• 7 seed stores across 7 cities (Mumbai, Chennai, Bengaluru, Delhi, Hyderabad, Pune, Kolkata) spanning ethnic wear, denim, and streetwear/footwear.

Seller Portal (/seller) — pick a demo store from the switcher (no auth in this build) and manage it:

• Dashboard — live stats (SKU count, low-stock alerts, plan, status).
• Products — full CRUD against the live catalog; changes appear on the storefront immediately.
• Theme Studio — color pickers + hero copy editor with a live preview, publishes straight to the store's theme.
• QR Marketing — generates a real, scannable QR code per placement (shop window, table, etc.), tagged for footfall attribution.
• Billing — Lite/Pro/Max plans, SKU usage, and GST invoices computed with real CGST/SGST-vs-IGST split logic.
• Reservations — incoming pickup holds for the store, with a live countdown, shopper contact info, and Mark Picked Up / Cancel actions.
• Onboarding — mocked GSTIN/PAN/UPI verification wizard.

What's mocked, on purpose: there's no real database (in-memory data, seeded from lib/seed-data.ts, resets on server restart), no real OTP/SMS delivery (shopper login, the seller "session", and the "seller notified" step for a new reservation are all local, no push/SMS service), and no real payments (plan changes don't call Razorpay/Cashfree — reservations were never meant to take payment; that happens at the store). Everything else — routing, filtering, CRUD, caching, cache invalidation, GST math — is real.

Repo map

Path	What it is
apps/web	The running app — consumer site + Seller Portal (Next.js)
docs/01-product-and-personas.md	Product scope, personas, core user journeys
docs/02-system-architecture.md	Target production architecture: microservices/serverless, infra, caching, CDN
docs/03-multi-tenant-storefront.md	Multi-tenancy model, theme engine, routing strategy
docs/04-monetization-and-billing.md	Plans, metering, Indian payment rails, GST invoicing (billing math is implemented; PSP calls are not)
docs/05-onboarding-and-compliance.md	GSTIN/PAN verification, bank/UPI setup, KYC flow (mocked in the app)
docs/06-api-surface.md	Target REST API surface for a real backend
database/schema.sql	Full PostgreSQL multi-tenant schema the in-memory data model mirrors
schemas/theme-config.schema.json	Storefront Theme Engine JSON Schema
schemas/theme-config.example.json

Target production stack (docs/02)

The docs describe where this goes in production — a real backend to swap in under the same API surface apps/web already calls:

• Frontend (B2C storefront + PWA): Next.js 14 (App Router, SSR + ISR), next-intl for i18n, Workbox for PWA/offline.
• Backend: Node.js/NestJS microservices (Catalog, Search, Tenant, Billing, Notification, Media) behind an API Gateway.
• Data: PostgreSQL (multi-tenant, row-level tenancy via store_id), Redis (cache + session + rate limit), Elasticsearch/OpenSearch (product search & geo queries), S3-compatible object storage + CDN.
• Search/Geo: PostGIS or Elasticsearch geo_point for radius search; India PIN code → lat/long reference table.
• Payments/Billing: Razorpay (primary) with Cashfree as secondary PSP — UPI AutoPay e-mandates, cards, net banking.
• Infra: Kubernetes (ap-south-1) or serverless (Lambda + SQS) for event-driven pieces.
• Observability: OpenTelemetry, Grafana/Loki/Prometheus.

Start with docs/02-system-architecture.md for the full picture, or just run apps/web to see it working today.