Skip to content

Point of Sale

Restaurant POS

A fast, offline-resilient point-of-sale suite covering front-of-house ordering, kitchen display, delivery aggregation and multi-branch reporting.

Client

Saffron Hospitality

Industry

Food & Beverage

Year

2024

Duration

7 months

Restaurant POS — Saffron Hospitality interface

Business Challenge

Peak-hour queues stalled whenever the internet dropped, delivery orders were re-typed from three tablets, and head office had no same-day visibility of branch performance.

Our Solution

We built an offline-first POS with local-first ordering and background sync, aggregated the delivery channels into one ticket stream, and gave head office a live multi-branch dashboard.

Engineering

Technology Stack

Chosen for longevity and operational simplicity, not novelty — every dependency here is one we still support today.

ReactTypeScriptIndexedDBNode.jsPostgreSQLStripe

6

Core technologies

7 months

Build duration

2024

Delivered

Inside the build

Project Features

Offline-first ordering

Every terminal keeps a local ledger and reconciles automatically once the link returns.

Kitchen display system

Course-aware ticket routing with prep timers and colour-coded urgency.

Delivery aggregation

Third-party channels land in one queue with unified menus and stock rules.

Table & floor plan

Drag-to-seat floor management with split bills and per-seat items.

Live branch reporting

Sales, voids and labour cost per branch, refreshed continuously.

Stock deduction

Recipe-level ingredient deduction ties menu sales back to inventory.

Development Process

Discovery

Two weeks of service-time observation across breakfast, lunch and dinner peaks.

Prototype

A tablet prototype tested by real servers during live shifts before committing.

Design

Thumb-reachable, high-contrast controls that work with wet hands and gloves.

Build

Local-first data layer with deterministic conflict resolution for order edits.

Hardware

Printer, cash drawer and card-terminal integration tested per branch model.

Rollout

Pilot branch, then a two-branch-per-week schedule with on-site training.

The Results

Average order time fell by more than a third and the chain expanded to new branches without adding back-office headcount.

38%

Faster order time

0

Sales lost offline

24

Branches deployed

3.1x

Reporting speed