Global knowledge analytics pioneer SAS has expanded its Viya platform with new agentic AI options. The developments embody AI assistants, agent infrastructure and acceleration instruments.
The updates construct on Viya’s present AI basis, which the corporate says can assist organisations transfer from remoted generative AI (GenAI) use circumstances to ruled, production-ready intelligence at scale.
The information was revealed throughout SAS Innovate 2026, the corporate’s world knowledge and AI convention, in Dallas final month (April).
SAS offers the next new options:
- SAS Viya Copilot – A household of AI assistants embedded throughout the analytics life cycle.
- SAS Viya Model Context Protocol (MCP) Server – Using the open MCP normal to show SAS Viya analytics and decisioning capabilities as instruments for AI brokers.
- SAS Agentic AI Accelerator – A curated framework for constructing, governing and deploying AI brokers inside SAS Viya.
“The role of human expertise in operationalising agentic AI is not diminished by automation; it’s elevated,” says Jared Peterson, SAS SVP of worldwide engineering. “With SAS Viya, organisations can pair copilots and agents with human judgment, trusted data and enterprise governance, so AI doesn’t just generate outputs but drives responsible, real-world decisions.”
Copilot for knowledge and AI duties
SAS Viya Copilot is a ruled, conversational AI assistant embedded instantly into the platform. The system is designed to work alongside human consultants inside production-grade analytics workflows. SAS says Viya Copilot can assist organisations analyse knowledge, construct fashions and make choices throughout the analytic life cycle whereas sustaining enterprise-grade safety and oversight.
SAS Viya Copilot integrates Microsoft Foundry to function inside analytics workflows reasonably than as a standalone chat interface. According to the corporate, Copilot helps knowledge scientists, builders and enterprise analysts generate, clarify and act on insights utilizing pure language throughout each coding environments and interactive visible interfaces.
SAS offers the next capabilities for Viya Copilot:
- General questions and solutions throughout core Viya functions. This consists of knowledge discovery, mannequin pipeline improvement, mannequin administration, choice intelligence, and setting administration.
- Code acceleration with AI-generated SAS and Python code, documentation and explanations.
- Model pipeline steerage with clever suggestions, subsequent steps and explainability.
- Conversational dashboarding with AI-driven knowledge enrichment, pure language dashboard creation and automatic insights.
- Visual investigation with AI-assisted search and AI-powered case and alert narratives.
Over time, says SAS, Viya Copilot knowledge and AI capabilities will prolong throughout knowledge administration, mannequin administration and AI infrastructure, enabling constant AI help from knowledge ingestion to deployment.
According to SAS, Viya Copilot is designed to execute domain-aware analytics duties grounded in trade knowledge fashions, terminology and finest practices, not like many horizontal AI assistants that rely totally on pretrained LLM information or generic instruments.
SAS offers the next trade Copilot capabilities:
- SAS Asset and Liability Management (ALM) Copilot, which guides analysts via state of affairs configuration, execution and interpretation for monetary threat workflows, translating natural-language inputs into ruled analytic fashions.
- SAS Health Clinical Data Discovery Copilot, which accelerates medical knowledge discovery and evaluation by guiding customers via knowledge exploration, cohort creation and high quality investigation utilizing pure language.
The firm plans to develop Viya Copilot for extra industries in 2026, together with monetary crime prevention in banking, and clever planning and provide chain optimisation for manufacturing.
From copilots to brokers
While SAS Viya Copilot can enhance how individuals work together with analytics, brokers permit these capabilities to compound by connecting insights, choices and actions throughout techniques at scale.
Beyond embedded AI assistants, SAS Viya can present the infrastructure and instruments organisations have to construct, join and govern AI brokers at scale. These capabilities, accessible to present Viya customers via GitHub, are designed to assist rework remoted GenAI use circumstances into enterprise-ready agentic AI that works throughout techniques, groups and enterprise processes.
SAS Viya Model Context Protocol (MCP) Server can present a standardised approach to expose SAS analytics, fashions and decisioning capabilities to exterior AI brokers. By utilizing SAS Viya MCP, organisations can embed SAS’ analytics, governance and area intelligence into their very own brokers from the LLM interface of their selection, reminiscent of Claude, with out duplicating logic or bypassing enterprise controls.
SAS Agentic AI Accelerator can allow groups throughout completely different talent ranges (no-code, low-code and builders) to design and operationalise AI brokers with governance and belief. The accelerator offers a curated assortment of code, elements, interfaces and finest practices to assist groups construct, govern, and deploy AI brokers inside SAS Viya, transferring from experimentation to manufacturing in a repeatable, ruled means.
SAS Retrieval Agent Manager (RAM) is a no-code answer constructed on the retrieval augmented era (RAG) framework that goals to streamline the method of remodeling uncooked unstructured knowledge into context-aware AI responses. While at present solely accessible as a standalone product, SAS plans to combine RAM capabilities into SAS Viya, additional enabling brokers and assistants to floor responses in an enterprise context for correct, trusted decisioning.
Together, says SAS, the updates place Viya as a ruled, enterprise-ready platform for agentic AI, supporting AI assistants and clever brokers that work collectively to compound worth and switch trusted insights into real-world choices at scale.
* To study extra about SAS’s Agentic AI merchandise, go to the web site here.
* Read Arthur Goldstuck’s protection of SAS Innovate 2026 right here.