SAS Viya continues to develop

SAS is one of the largest private firms in the world, running since 1976 and with its founder Jim Goodnight still running the company at a youthful eighty. SAS does over 3 billion dollars (USD) in annual revenue, with 90% of the Fortune 100 being SAS customers. It had a subscription only business model decades before that became fashionable in the software business, and spends about double the industry norm on R&D. It has an unusually high customer retention rate, and employee turnover of around a fifth the industry average.

The SAS platform, branded Viya®, has a series of core data management functions, with numerous industry specific applications e.g., fraud detection, sitting on top of the core platform. It has connectors to a vast range of data sources, from databases through to proprietary file systems and even to social media data. SAS Analytics aims to minimise unnecessary data movement, so for example has in-database execution in a number of platforms including Teradata, Spark and Hadoop. It has tools to help users set up data pipeline flow between sources and has links to other tools in this area, such as to Apache Airflow.

SAS has its own data management catalog and business glossary, allowing business rules to be set up and stored. It also supports the open source OpenMetadata solution for data exchange with 3^rd party sources such as other data catalogs. SAS has always been strong in data quality, though this is now embedded within its broader platform rather than sold separately as it once was under the Dataflux banner. SAS supports the full range of data quality functions including data profiling, match/merging, data cleansing and enrichment.

On the horizon are further developments such as “cognitive data quality”, where machine learning is brought to bear to improve automation in data quality. SAS continues to invest heavily in R&D, with a recent emphasis on artificial intelligence. For example, the Amsterdam UMC oncology centre uses AI within SAS Viya to scan clinical images alongside radiographers who are checking for cancer.

In another example, Honda uses AI capabilities within SAS to help it detect fraudulent car insurance claims, more than doubling their previous success rate at finding non-complaint claims.