Australia needs better cloud and data management capabilities

Cloud skills gap

Spira also touched on cloud skills, or the lack thereof, as companies moved to the cloud. “No one is going to be able to do something they have never done before. There’s a quantum jump between running your own datacentre to consuming computing through an API [application programming interface],” he said.

“And even if you bring in a consultant, you still need expertise on your own team. Data has a lifecycle, a system has a lifecycle and at some point it becomes legacy and will need to be rebuilt,” he said, adding that this is when the internal cloud skills become essential.

Internal skills are also critical to meet the shared responsibility that APRA requires. While only a segment of Australian business is regulated by APRA, its cloud paper provides useful guidance on the sorts of issues that cannot be simply be handed over to a third-party cloud supplier.

These include identity and access management, incident detection and response management, data loss prevention, vulnerability management, configuration management, encryption and key management. In addition, oversight of the scope and coverage of audits initiated by the service provider, as well as planning for business recovery in the event of a cloud failure or outage, are just as critical.

Data management snags

As companies move to multiple and hybrid cloud architectures, the challenges become magnified. According to Nigel Mendonca, ANZ country manager for Tableau, the days of organisations trying to hold everything in a single repository are over.

Mendonca said enterprises recognised the need to seek insights from multiple data sources – in-house, in the cloud and from third parties. The more experienced Australian organisations were comparable to the rest of the world in terms of orchestrating that data, he added.

However, Mendonca acknowledged that there was a big gap between leaders and laggards, and that Australia was probably a little behind international peers in terms of its move to democratise access to data regardless of where it was held.

He advised organisations to look at the data they have before moving it to the cloud. “What sorts of compliance obligations do you have? Are you subject to the GDPR [General Data Protection Regulation]? Once it’s classified, you can map out what data lives where.”

GorillaStack’s Spira called for cloud users to carefully review different cloud offerings to ensure they have one that meets the requirements of data that will be stored there.

The biggest mistakes he had seen were from organisations which had not considered tagging or indexing their data so they could find it again. “To tag manually is a pain, but you can automate it from the beginning,” he said.

Pooyan Asgari, chief data officer of Domain Group, said Australian enterprises were “not that mature” in terms of their approach to cloud management and use of data, ranking most businesses an “early stage two” out of a possible four stages of maturity.

He stressed the importance of cloud computing as it makes information easier to access and share, but noted that this often required businesses to add another layer to their infrastructure – using systems such as Snowflake to get a single view of data across multiple clouds.

Ramneek Gupta, managing director and co-head of venture investing at Citi Ventures, said at the recent Sibos conference in Sydney that one of the challenges facing many organisations worldwide was that while many aspired to having data lakes, many managed little more than “data puddles”.

Gupta said as data has a form of density or gravity, and to be able to extract insights from data, it was important for organisations to create a virtualised data collection and an abstraction layer called a data fabric on top of that. This would require a degree of cloud wrangling maturity that has yet to be developed by many Australian enterprises.