In the next three years, SAS will invest $1 billion in the development of new analytics solutions aimed at meeting the needs of certain industrial sectors such as banking, government, insurance, healthcare, commercial, manufacturing and energy. All solutions will run on SAS Viathe cloud-native artificial intelligence (AI) platform. An announcement that confirms the company’s strong commitment to AI-powered development.
Commitment which, moreover, had already been highlighted three years ago, when, by voice of his CEO Jim Goodnight had already announced a first investment in AI always of 1 billion dollars. Now comes this second tranche and the company proves to be more inclined than ever to want represent a point of reference in the field of analytics. He clearly reiterated this in Milan on the occasion of the Italian leg of the tour SAS Innovate which led the company’s management to meet partners and customers throughout Europe “to demonstrate what it means to innovate through the use of data”, as he stated Manuela Cerruti, Managing Director di SAS Italy e Regional Vice Presidentat the opening of the event.
Being data-driven is no longer enough
Brian Harris, Xxecutive VP e CTO di SAShe specified that “In today’s business world the race consists in identifying opportunities and making better decisions faster than competitors. But doing business is increasingly complex and, at the same time, information overload is far exceeding the human ability to derive value from it. It is no longer enough to be data-driven, you need to know how to exploit AI to scale in productivity and have a more effective decision-making process”.
A result that can be achieved because AI makes it available three important qualities: productivity, performance and credibility. “In AI, decisions are made based on models Harris said. Production is the only piece of evidence of productivity, everything else is a scientific experiment. And in today’s economic environment, no one has time for scientific experiments”.
The three qualities Harris cited are the foundation upon which the SAS Viya platform is builtwhich therefore “offer all the tools for effectively query the datamanage them transparently e create industrial solutions that increase productivity. “Our solution allows you to operate faster than the competition”, Harris pointed out.
Cost improvement
The second point Harris emphasized is the performance. “In the context of the cloud there are three qualities that are typically talked about: agility, resilience and cost improvement”, underlined the CTO of SAS. In the case of AI, the costs essentially depend on the resources and storage used. As far as resources are concerned, the costs concern the training of the models that is performed by leveraging machine learning and algorithms. “The R&D division of SAS has worked hard in this direction – said Harris – getting 16x faster learning algorithms than they were in 2021. But there’s more: a new independent external study has certified that our algorithms are on average 30x faster than commercial and open source alternatives”.
In more detail, as he specified Marinella Profi, Product Marketing Manager at SAS“the study was made by Future Group and the SAS Viya platform resulted 30 times faster, scalable and cost effective in 86% of cases compared to competitors. A result obtained by speeding up both the code and the computing, elements that act directly on productivity”.
On the side of storageHarris instead cited the partnership made with SingleStore which offers a cloud-native distributed SQL database and cloud portal that can integrate transactional and analytical workloads in the same engine. “Our two teams have been working together for a couple of years to create an offering that we believe represents the next generation of analytics architecture – Harris said visibly satisfied -. We have tried to obtain hitherto unthinkable benefits: in fact we have reduced storage costs by 60-80% thanks to excellent compression protocols some data. Also, we have increased the efficiency of the modeling process limiting the movement of data from the database to memory as much as possible”.
In this regard, Marinella Profi highlighted how there was a change of approach in cloud migration: “The most recent surveys show that for the first time the primary concern for CIOs is not security, but cost management”.
Responsible innovation
About the credibilityunderstood as confidence in the results obtained from processing using AIHarris specified that “any powerful technology carries risks and these risks must be understood. In the case of AI, it’s important to understand that is able to analyze (especially) the past and make predictions about the present or the future. However, if the data used for training is biased or deficient, the models created will give false results which could have unintended consequences. So it’s imperative that as a community of builders and consumers of technology, we ensure that decisions made by AI on behalf of people are explainable, transparent and fair. We call this responsible innovation”.
“Even with the cloud Harris added. most customers today spends 80% of time preparing data and 20% analyzing it. We want to overturn this relationship and we are to achieve that goal combining some of the most cutting-edge data management capabilities with generative AI”.
Da data-driven ad AI-driven
Harris pointed out that SAS wants to be sure to integrate generative AI in order to have responsible innovation. In that sense, they are patents have already been registered. “We already have some very powerful tools Harris concluded. ma what excites me is the potential we are working on. Economic volatility is changing customer behaviors at an unprecedented rate. As mentioned, it is no longer enough to be data-driven, today you have to be AI-driven. And we are going in that direction”.
Bring Your Own Language
Harris’s words were echoed by Shadi Shadin, VP Product Strategy di SAS. Shadin focused attention on the fact that when cloud-optimized algorithms are used for analytics, engineers can work much faster and the insights obtained allow more effective business decisions to be made. “The language used should no longer be a barrier to productivity – ha sostenuto Shadin –. Must be the concept of Bring Your Own Language: Whether you’re using Python, low code or no code, today you have to be able to take an AI model from development to production as fast as possible”.