Contino Builds a Modern Scalable Data Insights Platform on Azure for One of Australia’s Largest Universities
The customer is one of the largest universities in Australia. It is a world leading university for research, teaching and innovation.
The University was looking to better understand the lifecycle of student applications and enrolment patterns from multiple data sources to drive operational efficiency, increase admissions and improve the overall student experience.
Previously, they were faced with massive data knowledge gaps that prevented the admissions and external relations team from gaining a deeper understanding of the experience of domestic and international students throughout the application to enrolment lifecycle. They needed help accessing data insights that would reveal why students chose to enroll and why others applied but then enrolled elsewhere. It was also important to understand which recruiters and agents were bringing in applicants (or not) and those who had the best and worst conversion rates.
The admissions and external relationship team also needed access to clear and concise data in order to better understand the health of their student pipeline and to allocate their spend most appropriately. The university was keen to leverage machine learning here to establish better forecasting and gain insight into which faculties are on track to reach their target number of enrolments. They were also interested in updating their outdated reporting process in favour of modern dashboards with charts and drill down features.
The key challenge was to navigate the fragmented data sources as well as a legacy data warehouse, and utilise cloud technology to deliver a rapid and affordable outcome.
Contino delivered a highly scalable and actionable Data Insights platform that facilitates actionable business intelligence, enabling targeted enrolment, agent and retention strategies.
Specifically, Contino delivered the following:
- A modern data platform on Azure, specifically a fully event driven data pipeline that allows dashboards to be updated in a timely manner. The platform leverages Platform as a Service capabilities on the cloud which makes it highly available and scalable. Dashboards are able to be consumed with high concurrency
- The implementation of CI/CD tooling and Source Code Management (SCM) for the data platform leveraging Azure DevOps, allowing one-click deployments
- An uplift of the University Planning and Performance team skills and capabilities with Azure, CI/CD, SCM and data framework knowledge leveraging Contino’s dual delivery and upskill approach
- The configuration of on premises data extraction through Azure Data Factory Self-Hosted Integration Runtime
- Implementation of operational intelligence sending logs, configuring alerts and creating dashboards using Azure Monitor
- The development of Machine Learning models leveraging Azure Databricks
- The introduction of Self Service BI through Azure Analysis Services
- The implementation of powerful visualisation via Power BI to better understand student admission and enrolment behaviour
The Business Outcome
- The University was able to pull all their data together to understand the student lifecycle end-to-end in timeframes previously thought unthinkable, let alone achievable.
- Since moving all their data to a streamlined and centralised location, they have since been able to leverage this and create several other new dashboards using Power BI to support better decision making.
- The university is now modeling and visualising the student recruitment-enrolment pipeline to optimise marketing, streamline financial reporting and retire legacy infrastructure. Their new data dashboards report on a range of dimensions that were previously unavailable to help the faculty to understand their performance from new perspectives. It also allowed them to undertake targeted proactive activities like contacting potential students at a certain point in the pipeline to see if they required assistance with the enrolment process.
- Proving that constructing a serverless, fully-orchestrated data lake was not only possible, but resulted in lower ongoing costs and the ability to construct new data models from aggregated sources which gave real insights on prospective student populations.