//
(On stage at SAS Innovate, left to right: Bryan Harris, Roshan Shah and Bill Clifford.)
At the SAS Innovate 2025 conference in Orlando, analytics provider SAS (Cary, N.C.) unveiled a suite of technology enhancements and strategic initiatives designed to help enterprises navigate disruption and harness artificial intelligence at scale. The announcements, spanning digital twins, agentic AI, quantum computing, and Viya platform enhancements, reinforced SAS’s strategy to drive real-world business outcomes through data-driven decision-making.
Bryan Harris, CTO, SAS.
According to Bryan Harris, CTO, SAS, the company aims to help organizations respond to widespread uncertainty—economic, geopolitical, and technological—by giving them a “decision advantage.” That promise, he said, is being fulfilled through a portfolio of innovations that stretch across every layer of the platform.
Unreal Engine Meets Viya: Digital Twins in Production
Among the most prominent announcements was a new partnership with Epic Games (Cary, N.C.) to integrate SAS’s AI and analytics with the Unreal Engine to build operational digital twin environments. The collaboration is already in use at Georgia-Pacific.
William Collis, who leads gaming analytics and simulations at SAS, explained that game engines like Unreal can now simulate real-world systems with high fidelity, producing rich, complete datasets that analytics engines can use with precision. Combining this data with SAS’s analytics, he said, creates a powerful “digital laboratory” for business experimentation.
Roshan Shah, VP of AI and Digital Transformation, Georgia-Pacific.
Bill Clifford, VP of Unreal Engine at Epic Games, described how Unreal’s ability to ingest data from IoT devices, CAD models, and geospatial sources allows businesses to visualize and manipulate complex systems in real time. He emphasized that the integration also allows telemetry data from the simulated environment to flow into SAS models and even trigger state changes within the simulation itself—closing the loop between AI insight and system behavior.
Georgia-Pacific’s Roshan Shah, VP of AI and Digital Transformation, confirmed that a pilot project is already underway at one of the company’s manufacturing facilities. The goal, he said, is to optimize operations in a way that could deliver “double-digit returns.” While scaling this approach across more than 100 facilities presents a challenge, he characterized the opportunity as transformative.
The collaboration also has broad applicability beyond manufacturing. Clifford pointed to examples in automotive, aerospace, and entertainment, from NASA training modules to Unreal-generated commercial imagery and in-vehicle experiences for EVs.
SAS Positions Agentic AI for Enterprise Use
SAS also introduced its framework for “agentic AI”—a new class of AI applications that not only analyze data but make decisions and take actions within enterprise guardrails.
Marinela Profi Global Market Strategy Lead, AI & GenAI, SAS. (Click to enlarge.)
Marinela Profi, global marketing strategy lead for AI and generative AI at SAS, said the market is moving beyond passive models toward AI agents that are “context-aware and context-driven,” able to operate at scale with built-in governance and transparency. She emphasized that many fully autonomous or LLM-only solutions are already “falling short” in enterprise contexts because they lack determinism and human oversight.
SAS’s approach to agentic AI centers on three pillars, according to Profi:
- SAS Intelligent Decisioning, which provides a secure environment to build and orchestrate AI agents that combine LLMs, rule-based logic, and traditional models.
- SAS Viya Copilot, an AI productivity assistant integrated with Microsoft Azure, designed to help users generate code, build machine learning pipelines, and support domain-specific use cases such as clinical trial design and asset-liability management.
- A growing portfolio of SAS Agents, including new offerings for data engineering, such as the Data Mapper agent, which uses AI to align disparate data schemas and accelerate system integration projects.
Profi said SAS is focused on building agents that can not only take action but “defend their decisions”—a distinction that reflects the company’s long-standing emphasis on governance and explainability.
Quantum AI: From Experiment to Enterprise Integration
Amy Stout, who leads SAS’s quantum AI product strategy, shared findings from a SAS-led global survey of 500 business leaders. More than 60 percent of respondents said they are actively investing in or exploring quantum AI, but many also cited high costs, lack of understanding, and difficulty identifying practical use cases.
Amy Stout, Quantum AI Product Strategy, SAS.
To address those challenges, SAS is pursuing a three-part strategy:
- Invest in hybrid quantum-classical research to combine SAS’s analytic capabilities with quantum speedups.
- Develop product integrations that allow users to access quantum compute resources and apply them to real-world problems.
- Offer services and pilot programs, including recent work with Procter & Gamble to optimize complex processes using quantum algorithms from D-Wave.
Stout said the company is working closely with hardware providers including D-Wave, IBM, and Quera, and intends to offer quantum processing options (QPU) within Viya alongside traditional CPU and GPU compute environments. Harris added that, as with GPUs, Viya users will soon be able to select a QPU to accelerate specific modeling tasks—especially in optimization—without changing their workflows.
Jenn Chase, CMO, SAS.
Viya Platform: Making AI More Accessible
SAS CMO Jenn Chase introduced several enhancements to the Viya platform, describing them as part of the company’s effort to combine breakthrough innovation with foundational reliability. “In order to have a winning team, you need to do more than dunk,” she said. “You need to rebound, pass, and build culture.”
Jared Peterson, SVP, platform engineering, highlighted SAS Viya Essentials, a cloud-native, packaged version of Viya designed for small and mid-sized businesses. He said the offering responds to feedback that Viya could be “heavy” for organizations with limited infrastructure or deployment capacity.
Jared Peterson, SVP, Platform Engineering, SAS.
Peterson also previewed ongoing development of SAS Viya Workbench, a coding environment for developers that now supports Python, R, and SAS languages. Designed to lower the entry point for new users, Workbench allows customers to experiment with high-performance analytics in a flexible interface. It is available via AWS Marketplace and will soon enter public preview on Azure.
The company also showcased SAS Data Maker, a synthetic data engine made possible through SAS’s 2023 acquisition of U.K.-based startup Hazy. The product generates statistically representative synthetic tabular data and is targeted to users in regulated sectors like healthcare and financial services. Peterson said the tool is currently in private preview and will become generally available in Q3 2025.
AI Modeling and Talent Support
Udo Sglavo, VP, applied AI modeling, described how SAS is building models that are “lightweight, industry-aligned, and production-ready.” SAS Models are deployed in containers and can be quickly integrated into customer environments—allowing businesses to see value from day one, even without an internal data science team.
Udo Sglavo, VP, Applied AI Modeling, SAS.
Sglavo said that SAS’s modeling tools are particularly useful for companies seeking to augment their existing teams. For instance, a pharmaceutical firm might already have strong in-house models for clinical data but lack AI expertise for supply chain management—an area where SAS’s domain-specific models can fill the gap.
He also drew a distinction between models, which handle discrete tasks, and agents, which pursue broader goals. He explained that agents combine tools, including models, to automate complex processes while leaving room for human oversight when needed—a concept he described as the difference between “human in the loop” and “human out of the loop.”
AI Governance: From Risk Management to Competitive Advantage
In a session focused on responsible innovation, Reggie Townsend, VP, SAS Data Ethics Practice, described how the proliferation of AI tools is reshaping the demands of enterprise governance. From shadow AI use to security risks and data leakage, companies are facing new challenges in ensuring that AI use aligns with internal policies and external regulations.
Reggie Townsend, VP, SAS Data Ethics Practice.
Townsend noted that even companies not yet deploying AI are building governance strategies, calling it “a matter of oversight, operations, and organizational culture.” Citing SAS’s recent AI Governance Report and global trends, he said, “Governance is a catalyst for innovation; public trust has become the new currency for AI.”
“The organizations that thrive won’t simply be those that deploy AI first, but it’ll be those that deploy AI most responsibly,” Townsend added. “It’ll be those that recognize the strategic reality that governance is a constraint to innovation, as it’s always positioned, but it’s a necessary companion.
SAS’s governance offering includes a new AI Governance Map, available on sas.com, which helps organizations benchmark their readiness and develop tailored action plans. The company is also expanding its product portfolio to include a unified governance platform that will orchestrate and monitor AI models, systems, and agents throughout their lifecycles.
SAS Acquires Hazy to Expand Synthetic Data Capabilities
SAS Expands Viya Platform with New AI and Cloud Capabilities
Source link
