AI-Ready Business Workflow System
Cafe POS & Ordering System
A cafe and restaurant operations platform — QR ordering, kitchen workflow, payments, and an owner dashboard — structured for AI analytics from day one.
- Next.js
- Prisma
- SQLite
- TypeScript
- Tailwind
- Mobile-first UI
01
Problem
Built for Chiya Mandi, a working cafe. The operation needed ordering and workflow that run on the phones staff already own — no dedicated POS hardware, minimal training time, and reliable behaviour during peak hours. The owner also wanted answers paper can't give: what sells, when, and what's running out.
02
Solution
A single Next.js application serving every role in the cafe, with Prisma over SQLite (Postgres-ready) keeping the stack deployable on low-cost infrastructure.
- QR-code menu and ordering — no app install, staff approve each order
- Kitchen and counter displays with order states (queued → preparing → ready)
- Inventory with automatic stock deduction on paid orders
- Expenses, profit/loss analytics, and trend reports for the owner
03
Why this is an AI project
AI analytics fail in small businesses because the underlying data doesn't exist. This system creates it — timestamped orders, item-level sales, preparation times, stock movements, payment records — and the planned AI layer reads that structure directly.
- Natural-language daily business summaries
- Peak-hour and demand prediction for staffing
- Smarter inventory alerts from consumption patterns
- AI menu recommendations and combo suggestions
- WhatsApp ordering assistant on the QR flow
What makes it credible
- QR dine-in ordering with staff approval, kitchen and counter queues
- Inventory with automatic stock deduction, expenses, and P&L analytics
- Every transaction stored as structured, queryable data
- 27 automated tests across orders, analytics, and inventory
Best fit for
- Cafes, restaurants & small food businesses
- Operations still run on paper, WhatsApp, or Excel
- Owners who want dashboards and future AI analytics
Operating principles applied
- Workflow first, AI second
- Structured data before analytics
- Owner dashboard as the intelligence layer
- AI layer only after the core flow is reliable
Roadmap
- UPI / payment gateway integration
- Real-time kitchen and counter boards
- AI daily sales summaries and peak-hour insights
- Menu and inventory recommendations from sales patterns
- Multi-branch support
Want a system like this for your workflow? Discuss a workflow or see other systems.