GLFN

Client name: One Bird Golf

Industry: Entertainment

Target User: Golfers & Golf Course Managers

Project Type: Web Dashboard & IoT System

Website: GLFN.com.au

GLFN Mobile Golfer iPhone App

The challenge

The client’s vision was to gamify amateur golf by allowing players to place instant bets on achieving a hole-in-one at specific par-3 holes. The challenge was creating a frictionless, “zero-app-download” experience that seamlessly connected a user’s mobile phone to physical hardware on the course.

We needed to engineer a system where scanning a location-specific QR code would instantly launch a secure betting interface. The critical engineering hurdle was ensuring that the successful digital transaction (via Stripe) immediately triggered on-site IoT controllers and networked IP cameras (Nx-Witness system) to capture high-definition video of the subsequent shot, automatically associating the footage with the user’s account for instant replay delivery.

The solution

A web application with two faces: a mobile-first view for golfers to watch and share their shots, and a web dashboard giving administrators full visibility and control. We used an NX Witness server to configure and interact with the on-course cameras, with shot recordings stored in AWS S3. Customer payments run through Stripe, transactional emails (registration and password-reset flows) and campaigns through Klaviyo, and SMS triggers through TallBob.

Backend

The backend implementation is made-up of multiple microservices (most of the services listed here is contained within a separate Docker container, and orchestrated using docker-compose), such as:

  1. Django: This is the main application in the backend, this service is the one which handle the API request made by the Front-end, this service is also responsible for scheduling the async tasks
  2. Celery Worker: This is a worker’s service used to handle async tasks created/requested by the Django service; This service is required as handling long-running/resource intensive tasks (such as, creating video thumbnail, sending notification to user, etc…) can block the API request handling of the main Django service, therefore Django service deferred those tasks to this service. This service is also used to sync data from third-parties services, such as sync the states of the IP Cameras from Nx-Cloud.
  3. PostgreSQL: This is the main service responsible for storing the long-term data, generated by the user
  4. Redis: This service is used for storing short-term data and used as a queue for communication between the Django service and Celery Worker service; This service is also being used by Django service for caching some API responses data.
  5. Nx-Witness and Nx-Cloud: This is a third-party system which is used to manage the IP’s Cameras; The IP’s Camera connected to the IoT device will be managed via Nx-Witness; Then our Django service will use Nx-Cloud as a proxy to communicate directly to the remote Nx-Witness, to request the state of the IP’s cameras via the provided Nx-Witness API interface.
  6. TallBob: This is a third-party API, which is used to notify the users via SMS
  7. Klaviyo: This is a third-party API, which is used to notify the users via E-Mail
  8. Stripe: This is third-party API, which is used as our payment gateway, this service will handle all the required payment procedure
  9. Nginx: This is the gateway to our API/Microservice environments, this service will route the request made by user to the necessary microservices; This service also responsible for serving the user with the distributed Frontend files

The tech stack

Frontend:

  • React
  • Responsive website
  • Admin dashboard

Backend:

  • Django
  • REST framework

Camera integration:

  • NX Witness Server

Infrastructure:

  • Docker
  • AWS (EC2)
  • CI/CD piplines via Gitlab
  • AWS S3 for recordings

Integrations:

  • Stripe (payments)
  • Klavivo (email capaigns and transactional flows)
  • TallBob (SMS)

Front End

The Frontend is made using a combination of TypeScript and React, the project is also being leveraged using Material UI and react-form.

The system is able to sync data automatically from third-party systems, such as Nx-Witness, TallBob and Klaviyo. The system is also able to send notifications via SMS or email.

The engineering edge

This software solution had three significant hurdles that we were able to overcome:

The first was the periodic data sync from third-party services. It was interfering with the application UX, but we were able to overcome this hurdle by using a celery worker to handle the periodic data sync in the background so the UX was prioritised.

The second was the IoT devices provided by the client didn’t work as expected. The IP cameras weren’t able to be recognised by Nx-Witness. We were able to get around this issue by writing a script to configure the IP cameras as a DHCP service, which assigns the connected cameras to proper IP address for Nx-Witness to recognise.

The final hurdle was the high data consumption by the IP cameras to Nx-Cloud. We profiled the network traffic within the IoT devices and optimised our data-sync system in the backend to save un-necessary network traffic.

The result

The client was happy with the implementations, and the system has been generating income for the client.

Looking for a similar solution?

Are you looking for a similar software solution? Element Software can help your business operate more efficiently. Get in touch to discuss how we can help you.