CafePro – Multi-tenant Restaurant & Café Management SaaS (Next.js + MongoDB)

Self-hosted restaurant and café SaaS: touch POS with a table floor, recipe costing with live margins, rosters and a geofenced timeclock, HR, CRM and loyalty, invoices, expenses and a live P&L, plus a 3D website for every café.

by DEV3 TechnologiesVerified Author(0) 0 sales 2 viewsUpdated 24 Oct 2026
CafePro – Multi-tenant Restaurant & Café Management SaaS (Next.js + MongoDB) — preview 1
1 / 10

CafePro is a self-hostable, multi-tenant SaaS for cafés, coffee shops and restaurants, built by DEV3 Technologies on Next.js 15 (App Router + Server Actions) and MongoDB. One installation serves many businesses; each has its own outlets, staff, menu, customers, books and public website, and its own currency and time zone.

Try it: the live demo is read-only and has one-click logins for the Super Admin, the business Owner, a Manager, a Head Chef (a custom role) and a POS barista.

Menu, POS and real-time food cost

  • Recipe costing – ingredients are priced per pack; recipes have portions and a waste allowance and give a cost per portion that updates when an ingredient price changes, so every menu item shows its current margin.
  • Touch POS – a table floor per outlet, dine-in, takeaway and delivery tickets, held tickets, kitchen notes, role-based discount limits and printable receipts.
  • Payments & loyalty – split payments (cash with change, card, loyalty points, other), points earned and redeemed automatically, voids and manager-only refunds that reverse points and spend.
  • Cost snapshots – every order and invoice line keeps its price and recipe cost, so past reports never shift when the menu changes.
  • Menu engineering – each item classed as Star, Plowhorse, Puzzle or Dog from its popularity and margin.

People and rosters

  • Shift planner – drag-and-drop rosters with collision detection, short-rest and leave warnings, shift swaps and roster publishing by email.
  • Timeclock – IP allow-lists, GPS geofencing and paid/unpaid breaks, with attendance corrections and approvals.
  • HR – employee profiles with encrypted national and tax IDs, a document vault with expiry alerts, and leave policies with accrual, carry-over and up to three approval levels.
  • Recruiting – job postings on each café’s careers page, applications with résumé upload and a Kanban applicant pipeline that hires straight into HR.

Guests, money and reports

  • CRM – customers merged from the website, POS and inquiries, a loyalty ledger, and a lead pipeline fed by contact, catering and event forms with round-robin or by-source routing.
  • Invoices – catering, event and B2B invoices with inclusive or exclusive tax, PDFs emailed with a signed public link, payments and an automatic overdue status.
  • Expenses & vendors – receipts, categories and multi-tier approvals that escalate above each role’s limit.
  • Reports – P&L with period comparison on recipe-based or purchase-based COGS (with the variance between them), sales by day, hour, category, type and payment, cash reconciliation and the HR audit trail, all exportable to CSV, Excel or PDF.
  • Mail – per-business SMTP, Gmail app password or Gmail OAuth2, a retry queue with a delivery log and a template designer with {{Variables}}.
  • Task boards & campaigns – Kanban boards with WIP limits, checklists, attachments and @mentions; campaigns with milestones and progress from their tasks.

A website for every café

  • A platform marketing site plus a ready-made site for each business at /cafe/<slug>: live menu, banners, moderated reviews, careers and contact forms, edited in a block-based CMS with SEO overrides.
  • An animated 3D coffee hero (three.js / React Three Fiber) with a scroll-driven camera dive, in light and dark themes.

Optional AI (Google Gemini)

  • AI import of supplier price lists, menus, vendors, receipts and bills, customers, staff and rotas from CSV, Excel, PDF, Word, photos or pasted text into a review grid; commits are all-or-nothing and a batch can be rolled back. Spreadsheets and delimited text also import without an API key.
  • Assistants – recipe from a photo, receipt scanning, invoice lines from text, “explain this period” for the P&L, menu engineering advice, inquiry triage, applicant screening, a daily briefing and “Write with AI” in text fields and the CMS.
  • Numbers are always computed by the app; suggestions are saved only when a person saves them; each business can switch AI off, and every AI control is hidden without a key.

Built for many businesses on one platform

  • Fail-closed tenant isolation: 38 of the 52 Mongoose models are tenant-scoped, and a schema plugin throws on any query that lacks the business filter.
  • Dynamic RBAC: Admin, Manager and POS roles out of the box plus custom roles built from a 26-module × 6-action matrix (107 permissions), with expense-approval and discount limits.
  • Secure sign-in: two-step sign-in with an authenticator app or recovery codes, optional Google sign-in, invitations, password reset and revocable per-device sessions.
  • Money in integer minor units per business currency; POS and invoice totals share one pure function.
  • Super Admin console: provision businesses, impersonate for support and edit the platform website.

What's in the package

  • Full TypeScript source (71 pages, 51 API routes, 52 Mongoose models), .env.example and README.
  • 16 Vitest test files with 88 automated test cases, plus repo checks that every server action goes through a guarded wrapper.
  • HTML documentation: help file, installation and deployment guide (MongoDB Atlas, Cloudinary, AWS EC2, Vercel, Docker) and project and module documentation.
  • Demo data seeder (two outlets, a costed menu and 60 days of orders, expenses, invoices, rosters and HR data), Dockerfile, Docker Compose and PM2 configuration, and a read-only demo mode.

Requirements

  • Node.js 20.9 or newer and npm
  • MongoDB 5 or newer running as a replica set (MongoDB Atlas works; payments, loyalty and refunds use transactions)
  • Optional: Cloudinary (required on Vercel), SMTP or Gmail, a Google OAuth client and a Google Gemini API key for the AI features

More from this author

View all

Related in Enterprise SaaS Starters

View all