Qrious, a startup owned by New Zealand telco Spark, is working closely with Pivotal on the next generation of its as-a-service big data offering.
Qrious technology director David Wills said that version one, also based on Pivotal, was developed in 2013 as a proof of concept of an end-to-end data science platform because traditional methods of provisioning such systems were very costly.
Bringing a data science platform-as-a-service offering to New Zealand would make the technology available to organisations on a pay-as-you-go basis that provides the service to those that would otherwise not be able to afford it.
Version two will again use Pivotal's packaged Hadoop open-source software and Pivotal Cloud Foundry to allow customers to build their own apps. Software from Tableau is used for visualisation and to deliver dashboards.
"Version two of the platform will include support for cloud based multi-tenancy as well as a customer portal and self-service provisioning tools for developers and data science teams to access our data science services," Wills told ZDNet.
"In addition to these core features, we are continually enhancing the platform with new product releases and upgrades of our data science tool sets."
Wills said Qrious is an early adopter of Pivotal's packaged open-source software. Being there early will allow it to leverage new features and influence development, he said.
The downside of high-velocity open-source development is that components move at different speeds and don't always integrate, requiring additional development work. Using Pivotal's pre-packaged software alleviates that, he said.
At the same time as it is building the platform, Qrious is also focusing on building its data and information assets, Wills told a New Zealand Analytics Forum meeting on Tuesday. That will involve aggregating public data such as population statistics, weather information, location information, and more to make it available to customers.
Qrious is building a data services platform to enable data to be deposited and withdrawn via "ingestion engines" and APIs built using software from Apigee.
Wills said that this would help the company create a "big data ecosystem" across New Zealand.
Qrious also offers software-as-a-service business intelligence for customers to explore and analyse their own data and professional services to help them achieve actionable results.
Qrious' platforms are hosted in New Zealand in Spark's datacentres on EMC and Cisco hardware. Wills said this answers the data sovereignty regulations and concerns that some customers have.
However, he said that while Qrious does not plan to compete with Amazon or Microsoft on per-unit pricing, public cloud is an option for the future if sovereignty and regulation concerns ease.
Wills said organisational awareness of big data is growing, but implementations are being hampered by the same kinds of issues that IT projects have always been prone to: Lack of governance, funding, and sponsorship.
On top of that, big data projects are speculative in nature, and data science skills remain very hard to come by.
Agile methodologies answer many of these issues, he said, and organisations should look at Agile techniques such as quickly building inexpensive minimal viable products for testing.
"The fast eat the slow," he said. "Big data is no different."