SAS introduces lightweight packaged AI models advances industry solutions with packaged AI models

These tailored, lightweight AI models facilitate rapid deployment with low overhead, and are one of the more interesting solutions just announced by SAS.

LAS VEGAS – Data and AI solutions pioneer SAS kicked off this year’s version of their SAS Innovate here Wednesday with the introduction of a flurry of new product releases that allow the company to address a broader range of use cases, and which give customers considerably more choice in the solutions available to them.

“I am really excited where we are,” said Bryan Harris, Chief Technology Officer at SAS. “We have been talking about our flagship program for years, and now about integrations with Generative AI, but what we are doing now is putting all of our solutions on Viya to provide customers with persistent growth.”

While LLMs are getting much of the media attention today, Harris said that they are not the main focus of most mode consumption, but rather are a very small part of the modeling needs of real-world production deployments of AI and decision making for businesses.

“What we are announcing in this new offering are out of the box models which move beyond LLMs and can accelerate time to value,” he indicated. These models address more specific use cases such as fraud detection, supply chain optimization, entity management, document conversation and health care payment integrity.

All of these new industry-specific models are engineered for quick integration, enabling organizations to operationalize trustworthy AI technology and accelerate the realization of tangible benefits and trusted results.

“These lightweight AI models can be readily operationalized – to address AI challenges with unparalleled efficiency,” said Udo Sglavo, Vice President for AI and Analytics, SAS. “Models are the perfect complement to our existing solutions and SAS Viya platform offerings and cater to diverse business needs across various audiences, ensuring that innovation reaches every corner of our ecosystem, and empowers businesses to flourish in their distinctive environments.”

Sglavo indicated that these lightweight SAS models make AI accessible regardless of skill set. These assistants cater to nontechnical users, translating interactions into optimized workflows seamlessly and aid in faster planning decisions.

“This is part of our billion dollar investment we talked about,” Sglavo said. “They will target specific and unique industry concerns.We want you to become data science experts.”

The first one scheduled to reach market is an AI assistant for warehouse space optimization. Other SAS Models are expected to be generally available later this year.

“Expanding our platform for LLM to build GenAI cases for SAS Viya greatly expands our capabilities,” said Marinela Profi, Global AI Strategy Lead at SAS. “Generative AI alone will not solve the problems, and only two tasks on this specific use case was done by LLM. LLMs alone will not do it either. SAS Viya needs developers for this. It reduces customer costs of handling complaints by 8-15%, and also leads to a 30-40%  decrease in average resource time used.”

Reaction at the show from SAS channel partners was very positive.

“The packaged AI models give partners the ability to add our own technology, while at the same time being able to get some things outside a core competency from the market,” said Gavin Day, EVP, Office of the CEO, Partrners and Channels at SAS. Phase Two is actually them getting the models What’s important is the model, not where its running, which could range from SAS to open source.”