​Open Universities turns to Bluewolf to redefine student experience

Gaining a single view of each student has allowed Open Universities Australia to look to a future where machine learning can help them complete their degree in a more personalised way.

Open Universities Australia (OUA) is a higher education organisation formed and owned by seven of the country's large universities. Now 25 years old, OUA boasts 175 different programs and around 1,400 different classes from 12 universities, allowing students to gain tertiary training completely online through its Open Access program.

Although OUA is almost unrecognisable from the organisation that begun broadcasting lectures via television at 2am, OUA general manager Andy Sheats said somewhere along the way, the OUA business got "really complicated".

special feature

Digital Transformation: A CXO's Guide

Reimagining business for the digital age is the number-one priority for many of today's top executives. We offer practical advice and examples of how to do it right.

Read More

"The platform was built and the systems were built around a much more complex business model," Sheats told ZDNet.

Sheats joined OUA in 2017 with a focus immediately on simplifying its marketplace platform and the underlying IT. However, with 250 staff all based in Melbourne and focused on the student, OUA turned to IBM's Bluewolf to help it centralise everything.

"We use a lot of Salesforce; we use Sales Cloud, Marketing Cloud, Service Cloud, Advertising Cloud -- we do a lot of stuff," Sheats explained.

"We use quite a few different applications because the education tech landscape is a bit complex, we use quite a few different things, but we wanted to bring all of the base information into Salesforce so that we have one view of the student to figure out what to do with them."

Sheats said the project was aimed at getting all of OUA's data into one place so it can begin to use machine learning to help "home-in" on the student's experience, whether it's a piece of advice, recommendations, or identifying a problem or a potential problem faster and better than a human can.

"The first step in doing that is getting all the data in a place we can actually run machine learning -- we haven't done the whole machine learning thing yet, we've just launched in the last six months, but that's the path we're headed down," he explained.

Although OUA is yet to kick off machine learning initiatives, as a result of its digital transformation program, the organisation has streamlined the student onboarding process.

Previously, a student applying for a degree or changing degrees had to engage with OUA via a minimum of five phone calls over the course of, on average, 45 days due to the back and forth between the student, UWA, and the universities.

"We've got it down to two calls including the first one where we do the application and the last one where we tell them they're in and the average time is down to two weeks," Sheats explained. "So 44 days to 14 days, but more than that, handling time has dropped by about 80 percent which means that we can do more of them.

"It's so dramatically simpler than it was."

Things don't always go smoothly

The entire transformation took 10 months, but it wasn't without stress.

"This was a large-scale digital transformation that had multiple streams that needed to connect again and we weren't really just standing up an instance of Salesforce, we were working with several other components and partners to enable us to bring the threads together," Bluewolf Australia and New Zealand managing director Gavin Diamond said, adding also the process started with a four-week design phase that informed what was to be built, integrated, and tested.

To Sheats, it was the fastest nine months ever.

"I started in February and they came to me with this business case to do this project and I said, 'No way, don't do this project ... you know these things never work, the big bang massive re-platform everything in one go'," he explained.

"It turns out that the platform we had was denigrating so quickly beneath our feet that we had to move, the risk of staying was not acceptable, and so we did go into this very fast process because we actually had a business need to get off what we were doing -- that really drove the timeline."

Sheats told ZDNet that realistically OUA could have taken another six months, given the approach was so aggressive, but it was a small window the organisation had to squeeze into.

"Some periods you can launch in, some periods you can't, and for us, university begins for the year in February ... February is our biggest entry period, we needed to be live to support that," Sheats said.

"We launched much earlier than we would have otherwise and we did that knowing that the benefit of launching was greater than the pain of not launching, and we knew that we would have problems on launch -- and we accepted that."

December and January saw OUA and Bluewolf work to fix things in situ.

"I don't think anyone was happy with it, none of us were, but we needed to do it," Sheats added. "When things are hard that's when you know how teams actually work."

In spite of everything, Sheats said OUA had its first steady period of growth in five years, even though it launched a system halfway through the quarter.

"This was a fire in the building and we actually needed to change," Sheats added. "No one actually wanted to -- I didn't want to do the project, I would have been happy not to change, at least not this quickly, but it really needed to happen."

In 2015, OUA on-boarded La Trobe university.

"When we brought on La Trobe it took us eight months to get them loaded -- it was horrible, painful work. That was one of the things that highlighted that we needed to get better," Sheats added. "We just brought Australia National University on in two weeks."

RELATED COVERAGE