momius - Fotolia
LAS VEGAS -- As one of the leading vendors in the analytics market, software from SAS Institute Inc. has never been synonymous with openness. The company's size has allowed it to set the terms for how customers buy its software -- and, typically, that meant selling standalone software.
But times are changing. With the emergence of open source analytics tools, businesses have more options than ever, and those tools are increasingly the first choice for data scientists and business analysts.
The keystone is SAS' new Viya platform, announced in April at the company's Global Forum. Viya is an analytics platform that can be deployed in the cloud or on premises, and, in a shift for SAS, it is accessible to open source analytics tools. Developers can build applications that leverage the SAS analytics engine through REST APIs, and analysts can query data through the platform using open source languages, including Java, Python and Lua.
When it was initially announced, Viya had support for some traditional SAS applications, including Visual Analytics, Visual Statistics and Visual Investigator. At Analytics Experience, SAS' new flagship user conference, the company said Visual Data Mining and Machine Learning will be added to the platform later this year.
And there may be more to come. Fiona McNeill, SAS global product marketing manager, explained during a news briefing that the company is looking into consumption-based pricing models that would charge businesses based on the number of API calls to its analytics engine. This is the kind of thing other cloud-based machine learning platforms, like IBM Watson and Microsoft Azure, are doing, and it would be a departure from the traditional SAS model of selling a complete piece of software.
Fritz Lehmanchief customer officer, SAS
The current move toward embracing openness has been a long time coming. "We've been a bit slow, but once we made the decision, we've hit it hard," said Fritz Lehman, executive vice president and chief customer officer at SAS. "It's an acceptance that the rest of the world is doing open source, and we need to play."
The move isn't without risk for SAS. Some open source environments act as direct competitors to the company's software. Becoming too open risks inviting the wolf into the hen house.
But it's a necessary risk, explained Randy Guard, executive vice president and chief marketing officer at SAS. Customers want to use open source analytics tools, and they want to build applications in a service environment. To him, the shift helps keep the company's offerings relevant.
"In a good way, [our software] was focused, but in a bad way, it was in its own area," he said. "This expands it."
Is open source the best approach to business analytics projects?
Open source data visualization tools are taking off
A comparison of big data analytics software