Traditionally, analytics was perceived to be complicated, expensive and proprietary. That’s why the investment in, and usage of analytics was not very high. But organizations now acknowledge that analytics is a competitive differentiator.
BW CIO World met Marcio M. Dobal, Senior Region Vice President, SAS APAC at SAS India Forum 2017 in Mumbai. He spoke about a new platform called SAS Viya and how it will bring analytics into the hands of more people in the organization.
SAS Viya is a new, open analytics platform. It is a single, cloud-ready environment that serves everyone: from data scientists to business analysts, application developers to executives – with reliable, scalable, secure analytics management and governance essential for agile IT.
Q. What steps has SAS taken to make analytics simple for everyone?The approach that we took towards this was simpler visualisation tools. There is some simple analysis that you can make with these (visualisation) tools. A layman cannot do any programming to get the analysis. He needs a visual tool to do some forecasting. The simplicity that our tool presents helps inculcate a culture of analysis within the organisation. It’s the ease of use combined with the power of the tool. Top management already realises the importance of analytics.
We also help the customer do the analytics, in case they don’t have the tools. For this we have Results as a Service (RaaS).
Q. What is driving high performance analytics in business?Businesses want to shorten the analysis cycle. Until recently, most of the analytics was done on sampling of data, because it takes a long time to run a model. You do not run all the possible models with all the data. Now there is more data than ever, and there is also unstructured data.
High-performance analytics allows you to look at the entire set of data. The processing happens in memory, at the edge (where you collect the data), sometimes without storing the data. And it happens so fast because the processing is done in parallel. Because of these advances in processing, you can run more models and all possible scenarios.
For instance, in banking, they could look at more alternatives to do risk mitigation. That’s the question that high-performance analytics is trying to answer.
Q. The new thing this year at the SAS Forum is SAS Viya. Can you tell us about it and how enterprises can take advantage of this platform? Which verticals in particular?
Viya is a big shift for SAS. It is a new platform and it is also an extension of our existing platform – for on-premise software. It is a complete shift in how we deploy applications and where these are deployed. Viya has got some new algorithms in terms of neural networks and machine learning. It is a cloud-based platform and it offers elasticity to customers to start small and then scale up or down, depending on demand.
The other feature about SAS Viya is that it allows us to embrace open source. If you are an existing SAS customer (using on-premise software), and you need some of these specialised algorithms, then you can use Viya as an extension.
But if you are a new SAS customer then Viya is a good entry point because you can start small and then grow as your business scales up.
Q. What are the new opportunities that SAS Viya presents to you and how are you going to leverage those opportunities in India?Viya is friendly and visually appealing; it has smart routines. It opens doors for those in the analytics world to enter a flexible, capable and open environment. Generally, people have not made a lot of investment in analytics software.
We have been selling our solutions mainly to the finance sector. But this opens some new doors for us in government, media and manufacturing. Sensors are increasingly used in manufacturing, and there is data to be analysed. For instance, Bridgestone, a Japanese tyre manufacturing company, needs 1,600 data points to create a perfect tyre!
The promise that Viya is making is that it would let more people use analytics.
Q. Please explain how you have adopted open source platforms, particularly Hadoop.Open source brings in innovation. As a company, you want to be innovative in your solutions. And you need people who learned open source programming languages (like R and Python) to come up with their ideas and start developing quickly. But this can become anarchic. If you don’t control then some people would use the data and it could be transformed in different ways by different programmers. Do you have a library of routines? What about quality control? You could have security issues. So there has to be some level of governance.
The second aspect is that there could be some specialised routines that could be difficult to program in open source. For instance, SAS is good at certain routines like risk management or fraud mitigation. So we are saying, if you know Lunar, R or python then use your knowledge to write your code and embed those calls to SAS within your program, so that it is seamless to you.
So you can have the advantage of both, the open world and proprietary.
