How Snappit scaled its ordering app to handle 13.3 orders per second at 15K+ attendee events

The Software House scaled Snappit's remote ordering application to handle high-volume event traffic and prepared it for integration with Norway's leading mobile payment platform within a 2-month deadline.

When Norway's second-largest bank acquired Snappit's restaurant ordering app, they needed it transformed for mass events. Fast.

The existing technology worked fine for restaurants but couldn't handle thousands of simultaneous orders at venues with 15,000+ attendees. With just 2 months to launch and half the team needed, Snappit looked for experienced PHP developers who could move quickly.

Snappit


Snappit was a Norwegian technology company founded in 2011 that developed remote ordering solutions for restaurants. Customers ordered from their tables using the app, and orders went straight to the kitchen browser. Over 4,000 Norwegian restaurants adopted the platform before DnB Nord acquired it in 2017 for their Vipps payment ecosystem.

INDUSTRY

Events & Restaurant Technology

COUNTRY

Norway

SERVICE PROVIDED

Team Extension, Legacy Code Refactoring, Cloud Infrastructure

Challenge

In 2017, DnB Nord - Norway's second-largest bank- bought Snappit's ordering app to add a new service to Vipps, their mobile payment product. The goal was to enter the restaurant and event industries with a proven solution. Snappit had already proven itself with over 200 Norwegian restaurants using it, but the acquisition came with a catch: tight deadline and technology that wasn't ready.

The existing stack worked for restaurants but couldn't handle the volume that mass events would bring. Picture this: thousands of people at a concert or sports event all trying to order drinks at the same time. The app needed to process orders without breaking, integrate perfectly with Vipps, and be ready in just 2 months. Snappit needed to double their team immediately to meet this deadline.

Key challenges and requirements:

-> Outdated technology stack not built for high-volume, simultaneous orders

-> 2-month deadline to prepare the product for Vipps integration

-> Need to scale the app for venues with 15,000+ attendees

-> Refactor significant portions of legacy code from the restaurant app

-> Separate the ordering app from the larger restaurant platform

-> Build new admin panels and update the frontend to React

-> Deliver error-free performance for Vipps integration

Solution

The Software House provided a complete team extension within weeks -exactly what Snappit needed to meet their deadline. The approach was straightforward: bring in experienced PHP developers who could hit the ground running, plus DevOps engineers to handle the infrastructure challenges of scaling for mass events.

The TSH team rewrote the app as a React-based single-page application and developed new admin panels. We refactored 75% of the legacy code and modernized the entire tech stack -upgrading PHP, Symfony, Node, Bootstrap, and all frontend libraries. 

The ordering application was separated from the larger restaurant platform so VippsGo could function as its own entity, optimized specifically for event ordering.

Source: Vipps

Process


Team Composition:

• 3 Developers (PHP, React)

• 2 DevOps Engineers

• 1 Project Manager

Technology Stack:

• Backend: PHP 7.1, Symfony 3.4, Node 10

• Frontend: React, Bootstrap 4

• Infrastructure: AWS, Kubernetes, Grafana, ElasticSearch, Jenkins

Initial 2-Month Sprint

The first 2 months focused on making VippsGo ready for Vipps integration. The team rewrote the customer-facing app as a React-based SPA and developed new admin panels. The ordering application was separated from Snappit's broader restaurant platform.

Technical Modernization

The team refactored 75% of the legacy codebase that remained from Snappit's restaurant app. This included:

• Upgrading PHP to 7.1

• Migrating Symfony framework to 3.4

• Updating Node to version 10

• Modernizing Bootstrap to version 4

• Updating all frontend libraries

Cloud Infrastructure Optimization

The DevOps team made key changes to prepare VippsGo for high-volume traffic:

• Added indexes to database columns

• Implemented auto-scaling for peak demand periods

• Introduced stress testing to validate the system could withstand high request volumes

Ongoing Development (2+ Years)

After the initial launch, The Software House continued working with Snappit for over 2 years, focusing on scalability and maintenance.

Outcome

VippsGo successfully processed 13.3 orders per second at events with 15,000+ attendees, handling 24,000 users per minute during Norway's biggest charity broadcast.