Mediality Leverages Automation to Deliver Racing Data Faster on AWS

Formed after the separation in 2020 of Australian Associated Press (AAP), Mediality Pty Ltd offers diverse media and publishing solutions including the country’s premier press release distribution network. Its Mediality Racing division, formerly AAP Thoroughbred Information Services and then AAP Racing, has decades of experience delivering data on thoroughbred horses to clients such as wagering operators, horse owners, and individual punters. 

When Mediality was spun off from AAP, the business—and its subsidiaries such as Mediality Racing—inherited legacy data center and application architecture, with Windows-based workloads that were initially built nearly 40 years ago. Mediality recognized the need to modernize but lacked the investment capital to move towards an open-source architecture on the cloud. 

Specifically, Mediality Racing wanted to shift from bespoke Windows applications to web interfaces. Its primary database, built on Microsoft SQL, stores horse racing data back to the 1980s and is the core of the business. Mediality Racing supplies Australia’s major newspapers with information for form guides and has a long-standing reputation for data accuracy. Ensuring the integrity of its data during a planned migration was critical.

Mediality Racing had attempted a piecemeal approach to modernization, but this ended up adding rather than reducing workflow complexity. Meanwhile, several of its customers were asking for more modern data delivery formats, including application programming interfaces (APIs). The company had been delivering racing data via large XML files for many years. 

A company restructuring in 2021 provided the opportunity to streamline. Tim Mansour, technology initiatives manager at Mediality Racing, explains, “We decided to move forward with a greenfield approach to redesign our data platform to be cloud native, leaving the past behind and deploying modern technologies to boost workflow efficiency.” 

Mediality has highly skilled developers on staff, but most of their experience prior to this project was with the .NET framework, and they were struggling to keep up with the company's growth. To build upon its developers’ expertise, the business chose to work with Cevo, an Amazon Web Services (AWS) Partner. Mediality had other workloads on AWS and wanted to execute the data project on a trusted platform following cloud best practices. The company has an ongoing relationship with Cevo and valued its deep knowledge and experience in developing solutions for customers—including those in the racing industry—using AWS NoSQL and serverless technologies. 

Cevo quickly began helping Mediality Racing develop cloud-native data workflows, setting up an AWS Landing Zone and using AWS Amplify as a user-friendly development framework. Mansour says, “AWS Amplify has been incredibly useful because it allows us to deploy very quickly and easily, pushing code changes to new environments in about 10 minutes.” This faster deployment directly accelerates Mediality’s development process by cutting testing time in half, Mansour explains. AWS Amplify also detects if parts of code are broken and prevents deployment in such cases—thwarting potential errors in racing data due to breaks in code. 

After analyzing how data was flowing in and out of its core database, Cevo helped Mediality migrate from Microsoft SQL Server, a relational database hosted in a managed data center, to Amazon DocumentDB, a fully managed non-relational database service. 

Because racing workflows are cyclical and prone to spikes just before events, Cevo recommended that Mediality use a serverless, pay-per-use approach for data transfers. Mediality is now using AWS Lambda serverless code to check for and automatically retrieve input data as it’s updated. Data retrieval and ingestion are fully automated, event-driven processes. Many files that formerly required manual transfer are now sent immediately to customers, saving about 10–15 minutes per event. Previously, Mediality Racing’s account manager would spend at least 2 hours a day preparing and loading files for each race. “This project will finally allow our account manager to focus on business and product development,” explains Philip McLean, managing director at Mediality Racing. 

The increase in automation across all data processes has drastically improved operation-wide efficiency. Mediality also has higher visibility into workflows on the AWS Cloud, to see where further automation could be introduced. Its teams are currently putting the finishing touches on a public API, which will be a first for the business. 

With the API, Mediality expects to see even greater efficiencies in file transfer timelines. Currently, employees take 7–8 minutes to review updated racing files and validate the data before sending updates to customers. Luke Donnelley, operations manager at Mediality Racing, says, “We’re expecting to see a significant uptick—up to 5 minutes—in the speed that we can deliver data. Five minutes is very significant in the corporate online book-making industry in Australia, which has become ultra-competitive. It’s a race for information.”