The University of Birmingham Strives to Graduate to a Data-Centric Culture with Snowflake

Higher education institutions have a lot of plates to spin, and the University of Birmingham is no exception. Following a tough pandemic, the need to digitally transform had never been more pressing. The university needed to modernize its data capabilities to better serve staff, students and researchers—and it used the Snowflake Data Cloud to do it.

Consistently ranked in the world’s top 100 universities, the University of Birmingham attracts students, staff and researchers from all over the world to learn, work and collaborate. 

To achieve its goal of ranking within the top 50 in the next 10 years, the university wanted to become more data-centric to gain insights about how to improve student outcomes and create a more efficient and rewarding place to work and study. But being such a large and complex organization—with so many different departments and moving parts—meant this was a huge undertaking.

Without a single source of truth, version control and data management were complicated. “Multiple versions of the same data were being used and modeled in slightly different ways,” explained Julian Kobylarz, Business Intelligence Development Manager at the University of Birmingham. “And these data silos were creating a negative view of data. It was taking a lot of people’s focus away from running the university.”

The University of Birmingham’s data existed in multiple systems, including about 450 places across the IT department alone. “It was difficult to stitch together hundreds of systems to create a holistic view of university data that meets everybody’s needs,” said Kobylarz.

The university found that many staff members couldn’t access the granular data they needed to best serve students and colleagues. In many cases, members of the team were turning to their colleagues to get exported Excel spreadsheets they needed to do their jobs.

Beyond data accessibility issues, the university was also concerned about data growth. The University of Birmingham’s legacy business intelligence platform and data warehouse wasn’t scalable. It couldn’t process large data sets quickly, performance was limited, and staff spent vast amounts of time trying to manually pick valuable insights out of the data noise.

When the pandemic began, it brought the importance of digital transformation into sharp focus for the university’s IT Services team. “It was a big shock for the university,” said Tim Packwood, IT Innovation Manager at the University of Birmingham. “Like lots of higher education institutions, we’d been investing in our campus for a long time, and hadn’t necessarily put the same investment into digital transformation.”

The team needed a solution that could match the dynamic nature of the university both now and in the future. It looked at 48 different technology platforms, but ultimately decided on Snowflake’s Data Cloud.

“We reflected on our digital strategy and what we needed to underpin that,” said Kobylarz. “The granularity of reporting and differentiation between compute and storage really set Snowflake apart.” The automatic, elastic nature of the platform was also a key factor for his team—and Snowflake was cost effective, too. “It was the best option for us,” he added. “And on top of all that, there was a lot of evidence out there that Snowflake was easy to use as well.”

The university did a proof of concept with Snowflake on Learner Analytics,collecting and analyzing student data to improve student outcomes. Now it can identify areas for improvement and implement targeted interventions to support student success. “The proof of concept went really well,” said Kobylarz. “From talking about it to the first data assets that went out through the door, it took just three months once we had Snowflake on board.”

The team then had a quick start in March 2021 with Snowflake’s professional services team. “It was some of the best, most insightful training that I’ve been part of,” said Packwood. “All of the team have been fantastic in supporting us and making us think ahead and introducing us to other people and clients.”

“The training was tailored for us,” added Kobylarz. “The team at Snowflake actually reflected at the end of every day, spoke to us, and changed the sessions to suit our team’s needs.”

Now the team is using Snowflake in three other key areas. In research data management, the ability to store and manage large volumes of research applications, awards and outcome data in the Data Cloud means professional services teams can easily access and analyze data in a secure, scalable environment.

The university is also using Snowflake for financial analytics, to store and analyze financial data—meaning the team can make data-driven decisions about budgeting and resource allocation. And by using Snowflake to analyze operational data, such as facilities management data, it can improve operational analytics and efficiency.

“One thing the platform is giving us is an immense amount of agility,” said Kobylarz. “Whenever somebody has a project they’re ready to start, we’re ready.”

Adopting the Snowflake Data Cloud has given the University of Birmingham’s IT Services team new capabilities that wouldn’t have been possible on its legacy infrastructure, including IoT, regulatory reporting, visualizations, and secure, fast data sharing. “We have the capability to crunch almost infinite data,” said Kobylarz. “We’ve got all the access controls in place and we can consume data in a very quick way. Whether it’s structured, unstructured, semi-structured, we’re ready to go.”

“The analytics we can do now just wouldn’t have been possible on our old infrastructure,” explained Packwood. “The capability gap has been closed.” 

“Working at a university, requests go from ‘Can you upload this Excel spreadsheet?’ to ‘I’ve got a billion things coming through every minute on an Internet of Things project—will you be able to handle that?’ Now, with Snowflake, I can say yes,” Kobylarz said.

Adopting Snowflake has had a huge impact on how the IT Services department can support people across the university. For example, Learner Analytics alone is helping academics find patterns in what they are getting right, identify what is working well, and roll that out across the university so all students can benefit.

“We can now look at trends to see, for example, if one professor is getting students to watch more videos than another. And if their students are getting say 10% higher grades, maybe students are consuming video better than they can consume a textbook,” Kobylarz explained.

Kobylarz, Packwood and the team are planning to roll out the platform in the future so non-IT professionals across the university can benefit as well. They plan to get people the right training and access to data to make it a key part of university culture.

And to improve strategic decision-making, the team plans to establish efficient and scalable self-service enterprise data and analytics platforms to continuously improve data quality, value and optimization.

“There’s no digital transformation without data,” said Packwood. “And Snowflake is an integral part of that. Now it’s about making sure the university can develop its culture around data. The ability to imprint and map our complex organization to Snowflake is the real, true power of the platform. And I hope in the next few years we really start to see it as the center of an ecosystem.”