Say hello to ‘Helios’ – A New Data Platform for Compass Group Australia


New data platform for Compass Group Australia 

About Compass Group Australia (CGA)

CGA is the Australian operation of Compass Group, one of the largest food and support services companies globally who are committed to delivering superior service in the most efficient way. They touch the lives of hundreds of thousands of Australiansserving over 70 million meals per year across 600 sites supported by their 11,000 Australian employees. Their services range from aged care and hospitals to corporate events, high schools, airlines, and remote mining sites.

The Challenge

An all-in migration to AWS was conducted by CGA in 2016, however many of the legacy systems remained disjointed. This resulted in data silos across the business and made it hard to analyse their data. Business users spend a large amount of their time manually assembling data and preparing reports in spreadsheets. 

With limited data and BI infrastructure, creating new reports was time-consuming and expensive. Key business metrics often had multiple versions of the truth, making it difficult to trust the analyses and the subsequent derived business decisions. 

The Solution

In 2021, CGA formed a bold new vision for data analytics and engaged Intellify to build their data platform. The platform was dubbed as Helios, named after the original Greek god of light and the keeper of oaths. Helios signals a commitment to provide trustworthy data products with the power to shed light on the darkest corners of CGAs data landscape. 

Intellify implemented an Amazon Redshift data warehouse, built reliable pipelines extracting and loading data from 10 data sources, and began producing several data marts for key business areas. 

 The central pillars of Helios are: 

  • Amazon Redshift RA3 nodes as the sole data warehouse. This allows for clear separation of storage and compute which facilitates an Extract Load Transform (ELT) pattern as opposed to the more traditional ETL pattern. 
  • Amazon Redshift cross account data sharing to allow data to be ingested once and then being shared between multiple dev/test environments. 
  • Amazon Managed Workflows for Apache Airflow to orchestrate the data and ML pipelines. 
  • AWS CodeBuild for CI/CD automating the testing and deployment of new ELT pipelines. 
  • Fivetran for data Extract and Load from various sources 
  • dbt for data Transformations and data lineage management. 
  • AWS Lambda and API Gateway for event-triggered data ingestion from custom data sources. 

The entire Helios platform is built on a well-governed AWS environment comprising multiple isolated hardened accounts.

Results & Benefits

The Helios data platform and the choice of Amazon Redshift RA3 drove a step change in CGA’s analytics capabilities. 

The newly created data application in the form of data marts became the trusted source of data for BI reporting giving back the confidence in data to the decision-makers. The data applications are designed for self-service reporting, allowing the business to quickly assemble reports, significantly reducing the reliance on spreadsheets. Manual reporting in spreadsheets was replaced with automated reporting from these data applications, saving 100’s of hours a month for the Compass team and ultimately allowing them to spend their time on high-impact activities. 

CGA was able to establish an in-house data analytics team to evolve and grow the Helios platform as new business requirements came in. Given the scarcity of data talent, the platform was designed for ease of use and low overhead allowing new data to be onboarded, transformations built, tested, and deployed within hours. 

Apart from delivering data applications, Helios has allowed CGA and Intellify to experiment and deliver more advanced machine learning projects for CGA. Some of these experiments turned into highly successful deployments such as forecasting customer site attendance and monitoring the nutrition intake of aged care customers.  

Helios lowered the bar for data experimentations and, we are thrilled to continue building many more exciting analytics and AI/ML applications in 2022.