SAS Viya helps businesses put GenAI models to work faster through robust LLM orchestration.

SAS, the global leader in analytics software and services, has announced that organisations leveraging the SAS® Viya® data and AI platform have been able to realise significant productivity and efficiency gains. Through the platform’s large language model (LLM) orchestration strength, SAS Viya is on average 30 times faster than all competitors tested across all test cases, while reducing operational costs by 86%.

"At the heart of SAS' core principle is universal access to data querying and advanced analytics across all stages of the data and analytics process," said Deepak Ramanathan, Vice President, Global Technology Practice, SAS. 

"That's why we're buoyed by initiatives like the Singapore National AI Strategy 2.0, as it not only recognises the pivotal role GenAI plays in this democratisation effort, but also the importance of bridging the gap between generative AI preparedness and execution," he added.

The GenAI capabilities within Viya represent a significant advancement for organisations aiming to optimise their processes and enhance decision-making through innovative AI solutions. Viya seamlessly integrates external GenAI models with existing business frameworks, facilitating end-to-end utilisation of LLMs across various enterprise use cases. These integrated capabilities are now accessible through SAS Viya, offering organisations a comprehensive solution to leverage the power of AI in their operations. In addition to seamless integration of GenAI models into decision workflows and existing business processes, Viya's quantitative decisioning capabilities also enable successful GenAI reasoning, which makes decision-making more precise and impactful.

Meanwhile, Viya's robust data quality measures that include minimisation, anonymisation, and encryption. At the same time, Viya facilitates the creation of trustworthy and explainable results by enabling data experts to preprocess data and explain generated outputs effectively, thereby reducing errors and improving transparency. Governance is also enhanced by Viya's built-in tools which create workflows that validate LLM lifecycles, including model risk management.

Organisations will also be able to tap into SAS' personalised assistant, Viya Copilot, to streamline analytical, business, and industry-specific tasks through an array of tools for code generation, data cleansing, exploration, marketing planning, journey design, and knowledge gap analysis.

With organisations exploring GenAI, SAS prioritises identifying industry-driven and ethically applied use cases. To that end, Deepak stressed that 2024 will see SAS continue to broaden its trustworthy GenAI footprint with the introduction of its own synthetic data generator – SAS Data Maker – and by offering industry-specific GenAI assistants. With the synthetic data generation tool, organisations can address limited data challenges by generating synthetic tabular data that statistically represents original training data without compromising sensitive information. This enables them to innovate faster by combining trusted data augmentation and generation algorithms in one seamless tool.

SAS is also continuing to integrate GenAI capabilities into its flagship MarTech solution, SAS Customer Intelligence 360. This integration enables marketers to elevate the customer experience by leveraging GenAI assistance for marketing planning, journey design, and content development. Additionally, three new capabilities have been introduced within SAS Customer Intelligence 360, enabling marketers to build recommended audiences, interpret audience data through a chat interface, and receive GenAI suggestions for email subject lines.