CBA CIO, David Whiteing, says he is interested in embracing a range of technologies more commonly associated with businesses that have a purely digital heritage rather than the complex technology environment of a bank.
The CIO is interested in technologies, including Docker, Apache Mesos, Kafka, and Cassandra, employed by many digital businesses, but seeking to apply them in the context where accuracy is vital.
“There's no question that, if you're sitting in the core banking space, where you've got to optimize for accuracy and reliability,” he says
“It's different to, if you're in the digital space, where you've got to optimise for customer feature delivery. They have very different cadences; they've got very different challenges.”
Whiteing made the comments as CBA unveiled a major revamp of one of its key data centres.
The CIO says that his team was eyeing opportunities for increased infrastructure automation and a “continuous, DevOps mindset” that will see increasing self-service by developers.
“The thing that I relentlessly worry about is: Are developers spending their time developing or are they spending their time coordinating with other teams and running around asking for business case funding? -- doing all the non-value added stuff rather than writing code.”
Whiteing says he wants to create the “maximum opportunity for them to write code”.
“Then if we get that right, we're going to deliver more features to customers more quickly.”
He doesn’t pretend the shift towards a more DevOps-style approach will be easy; it requires “a big cultural change” the CIO says.
“In our team of 8500 people, there's only a handful, there's probably just over 1000 of them, that are steeped in that way and we need to take the rest of them into that space,” Whiteing says.
His vision is to reorganise that workforce into what he describes as “platform teams”.
“Each of those platform teams will then become the determinant of which technologies they want to adopt, how fast they want to go,” the CIO says. Those teams will become consumers of infrastructure services, in a manner akin to teams at Google and Facebook.Read more:Embracing enterprise DevOps
“That team can then optimise for that automation and that self-service,” Whiteing says. “A platform team can then set their priorities and pace and drive things that are appropriate for them.”