WorkSmart Hisab – Canteen POS & Meal ERP Platform~5 min read
Flutter & Cloud ERPActive DevelopmentProprietaryTechnical lead

Smart Hisab – Canteen POS & Meal ERP Platform

High-speed, offline-first canteen POS and institutional meal ERP for Android — built with Flutter and Supabase PostgreSQL.

Role:Lead Mobile & Backend Engineer (Client Project)
Timeline:2026 — Present
FlutterSupabaseDartPostgreSQLAndroid

Architected and developing an offline-first institutional canteen POS and meal ERP platform for a dedicated client using Flutter and Supabase PostgreSQL [X]. Replaced paper meal tokens and chaotic shift registers with automated customer AR accounts and digital shift cashbooks [Y], enabling sub-second counter transactions, zero-variance cashier handovers, and atomic ledger reconciliation during peak lunch hours [Z].

My contribution

Technical lead

Architected and developing an offline-first institutional canteen POS and meal ERP platform for a dedicated client using Flutter and Supabase PostgreSQL [X]. Replaced paper meal tokens and chaotic shift registers with automated customer AR accounts and digital shift cashbooks [Y], enabling sub-second counter transactions, zero-variance cashier handovers, and atomic ledger reconciliation during peak lunch hours [Z].

What I owned

  • Guarantee sub-second counter throughput during lunch rush without internet dependency
  • Maintain financial double-entry integrity between customer payments, cashier cashbooks, and vendor food purchases
  • Support multiple user clearance levels (Superadmin, Cashier, Kitchen Manager, Auditor)

Product context

Commercial Canteens, University Dining Halls, Factory Cafeterias & Institutional Caterers

Architected and developing an offline-first institutional canteen POS and meal ERP platform for a dedicated client using Flutter and Supabase PostgreSQL [X]. Replaced paper meal tokens and chaotic shift registers with automated customer AR accounts and digital shift cashbooks [Y], enabling sub-second counter transactions, zero-variance cashier handovers, and atomic ledger reconciliation during peak lunch hours [Z].

Core challenge

Institutional canteens, corporate cafeterias, and educational dining facilities operate under extreme peak burst loads: hundreds of diners arrive within a 45-minute lunch window. Reliance on paper tokens, manual tally books, or cloud-dependent web systems results in catastrophic counter queues during Wi-Fi drops, misplaced meal receipts, cash drawer discrepancies, and uncollected customer debt.

Value proposition

Peak-Hour Counter Congestion — Web apps freeze or lag when 300+ students or workers queue for meals simultaneously on spotty cafeteria Wi-Fi.

Offline-First Instant Dispatch — Flutter SQLite engine issues meal tokens in <1.2 seconds with instant Bluetooth thermal receipt printing.

Cash Discrepancies & Drawer Leaks — Multiple cashiers change shifts without structured drawer counts, leading to untracked register shortages.

Atomic Shift Cashbook — Enforces mandatory opening float declaration, mid-day cash drops, and closing drawer audit before handover.

Uncollected Meal Credit — Manual credit ledgers lead to disputes over diner meal tabs and unpaid month-end balances.

Real-Time Meal AR & Wallets — Digital diner balances, monthly pre-paid credit limits, and automated payroll deduction exports.

Proof

Impact and product screens

My results

<1.2sCounter CheckoutRapid meal voucher issuance via offline SQLite queue
0.0%Shift VarianceStrict cash drawer open/close reconciliation
500+ MealsPeak CapacityHandles concentrated rush without network latency

Product outcomes

100% RLSData IntegrityPostgreSQL tenant and role-based clearance

Click any preview to open full size. Use arrow keys to browse.

Build

Stack and engineering choices

Technology

Frontend
  • Flutter (v3+)
  • Dart
  • Offline-First State Management
  • Material Design 3 UI Engine
  • Haptic & Sound Feedback
Mobile & Hardware
  • Android Native Build (APK)
  • Local SQLite Cache & Queue
  • ESC/POS Bluetooth Thermal Printing
  • Barcode / QR Token Scanner
Backend
  • Supabase (PostgreSQL 15+)
  • Row-Level Security (RLS)
  • Atomic RPC Transaction Functions
  • Realtime Sync Channels
Database
  • PostgreSQL Relational Schema
  • Customer Meal Accounts Receivable (AR)
  • Business Day Shifts & Cash Drawer Vault
  • Vendor Inventory Accounts Payable (AP)
  • Staff Meal Payroll Deductions
DevOps & Infra
  • Supabase CLI & Local Docker Engine
  • Automated Migration Replay
  • Android Production Keystore Pipeline
  • Git & GitHub

Key decisions

Institutional canteens, corporate cafeterias, and educational dining facilities operate under extreme peak burst loads: hundreds of diners arrive within a 45-minute lunch window. Reliance on paper tokens, manual tally books, or cloud-dependent web systems results in catastrophic counter queues during Wi-Fi drops, misplaced meal receipts, cash drawer discrepancies, and uncollected customer debt.

Guarantee sub-second counter throughput during lunch rush without internet dependency

Decision: Built an offline-first local mutation store in SQLite with optimistic UI updates and background batch sync

Outcome: Cashiers complete orders continuously without waiting for remote server round-trips

Maintain financial double-entry integrity between customer payments, cashier cashbooks, and vendor food purchases

Decision: Implemented atomic PostgreSQL stored procedures (RPCs) in Supabase with explicit debit/credit balancing

Outcome: Eliminates partial ledger updates and guarantees zero discrepancy between drawer balances and meal orders

Support multiple user clearance levels (Superadmin, Cashier, Kitchen Manager, Auditor)

Decision: Applied Supabase Row-Level Security (RLS) policies tied directly to user roles and assigned shift IDs

Outcome: Restricts cashier visibility to their active shift while allowing management real-time cross-location audit

Trust

Security and roadmap

Security & reliability

  • Row-Level Security (RLS) enforcing strict tenant and role isolation on all PostgreSQL tables
  • Atomic SQL transactions ensuring that meal order recording and wallet balance deductions never desynchronize
  • Encrypted local SQLite storage safeguarding offline transaction buffers on Android hardware
  • Cryptographically signed session tokens with automatic background token refresh

Future roadmap

  • RFID / NFC smart card tap-to-pay reader integration for contactless 0.5s counter clearance
  • Automated kitchen display system (KDS) for real-time prep station order dispatch
  • Weekly nutritional tracking and dynamic meal subsidy calculation for corporate cafeterias
Progressive Disclosure

Architectural Deep-Dive, Schemas & Workflows

Jump:

At a Glance

AspectImplementation Details
PlatformNative Android APK built with Flutter & Dart
Backend & CloudSupabase Cloud (PostgreSQL 15+, Auth, Realtime, Storage)
Offline ArchitectureLocal SQLite write-ahead transaction buffer with automatic conflict resolution
Hardware IntegrationsBluetooth ESC/POS 58mm/80mm thermal receipt printers, 2D barcode scanner
Core ModulesCashier Shifts, Counter POS, Customer Meal AR, Vendor AP, Staff Payroll Logs

The Operational Challenge

Institutional food services operate differently from standard retail shops. Rather than browsing items, diners want rapid, frictionless service:

  1. Burst Volume: 80% of daily transactions occur within two concentrated 45-minute periods (breakfast and lunch).
  2. Hybrid Settlement: Diners pay with cash, digital wallets, or institutional credit accounts (deducted from monthly salaries).
  3. Cashier Accountability: Canteens operate across multiple shifts with rotating operators; without strict cash drawer float tracking, reconciliation at end-of-day is impossible.
  4. Kitchen Coordination: Counters need fast item tallying to alert cooks when high-demand dishes (rice, curries, snacks) are running low.
flowchart TD
    subgraph Client ["Flutter Mobile Client (Android)"]
        UI["Counter POS & Touch Grid"] --> Queue["Local SQLite Queue"]
        Queue --> Printer["Bluetooth Thermal Printer"]
    end
    
    subgraph Sync ["Background Sync Engine"]
        Queue -->|Sync mutations when online| RPC["Supabase Atomic RPCs"]
    end
    
    subgraph Cloud ["Supabase PostgreSQL 15+"]
        RPC --> Shifts["Shift Cashbooks"]
        RPC --> Ledgers["Customer AR & Wallets"]
        RPC --> Stock["Vendor AP & Raw Stock"]
    end

Key Functional Modules

1. Shift & Cash Drawer Control

Before processing any transactions, cashiers must open an active business shift by declaring the starting physical cash float. Every incoming cash payment, customer cash-in, and expense payout is tracked against that active shift. At handover, the app calculates expected versus physical cash, logging variances for management review.

2. Customer Meal AR & Institutional Accounts

Regular diners, faculty, or corporate employees maintain digital meal accounts with configurable credit ceilings. The system allows post-paid dining with automatic ledger updates, generating end-of-month statements for direct payroll deduction or cash settlement.

3. Rapid Counter POS & Hardware Integration

The UI is optimized for large touch targets and minimum taps. Counter staff can add items, select payment method (Cash, Digital Wallet, or Account Credit), and print an itemized kitchen slip in less than two seconds.

Case closed

More work.

Email meAll work →