Restaurant Analytics Dashboard
At Mobikon, we built a full-stack CRM platform for restaurant customer engagement — serving restaurant chains across Singapore, India and the United Kingdom. I worked on both the Scala/Play Framework backend and the AngularJS manager dashboard that restaurant operators used daily.
The platform combined feedback collection, loyalty programs, marketing campaigns, and order management into a unified analytics dashboard. Elasticsearch powered aggregations across millions of records, while WebSocket connections enabled live updates — managers could see a customer's feedback appear on their dashboard in real-time.
Interactive Demo
A replica of the mEngage manager dashboard. Hover over the orange sidebar to expand navigation. Widgets display sample restaurant metrics with live updates every 3 seconds via the same WebSocket pattern used in production.
Key Features
Live Analytics Dashboard
Customer feedback, orders, bookings, and loyalty data from restaurant locations are aggregated into a dashboard with customizable chart widgets — giving managers visibility into what's happening at every outlet.
Smart Customer Campaigns
Segment customers by demographics, order history, and satisfaction scores, then reach them with targeted email and SMS campaigns — all with built-in budget controls.
Automated Alerts & Reports
Managers are notified instantly when negative feedback comes in. Daily digests, analytics exports, and industry benchmarking reports are generated and delivered automatically on schedule.
Technical Highlights
Shared Component Library at Scale
A shared AngularJS library was developed and maintained as a Git submodule across multiple production applications. The real-time analytics dashboard used WebSocket connections to support high-frequency data updates, with independently rendered widgets consuming Elasticsearch aggregations and displaying live AmCharts visualizations.
Percolation Alerts
Elasticsearch percolation enables inverted/reverse search — managers pre-register alert queries (e.g., 'feedback rating < 3 at outlet X'), and every incoming document is automatically tested against all registered queries, triggering instant push notifications on match.