The rise of agentic AI providers is much outstripping the earlier generations of predictive and generative AI that had us momentarily gripped not so very way back, because the enterprise expertise {industry} regarded to use these capabilities to trendy stacks and enterprise workflows.
Pre-agentic capabilities nonetheless exist, in fact, though many are actually subsumed into the substrate of wider and extra advanced providers which can be spearheaded with agentic entrance ends – and people are entrance ends being engineered to allow robotic work buddies to sit down alongside their human counterparts in more and more automated roles from excessive finance to the manufacturing facility ground.
Sector-specific AI, particularly
What could also be lacking then is the bridge to not simply utilized intelligence at a basic {industry} degree, however extra industry-specific AI providers the place brokers and fashions are constructed and aligned to extra devoted datasets and extra outlined (typically deterministic, some extra random) operational enterprise outcomes.
Knowledge and AI platform firm SAS says that it’s additionally a query of finances and time constraints, all of which might result in a “transfer quick and break issues” mentality that lets vital governance go by the wayside. That fail quick & fail typically mindset works properly with Agile software program software growth in its purest sense, however within the fragile quantum state that’s generative AI, it’s not all the time appropriate.
On a mission to supply some solutions on this house, SAS says it’s working to equip clients with {industry} accelerators with AI brokers and fashions to unravel each {industry}’s hardest challenges.
Provide and ops planning (S&OP)
Amongst its latest providers is SAS Provide Chain Agent, a expertise designed to streamline provide and operations planning (S&OP), a course of retailers and producers use to handle provide chains as markets and materials availability ebb and circulation.
In keeping with professionals who analyse this house, S&OP is a “multi-day taxing course of” (taxing as in exhausting work, not as in working to handle nation and state levy prices for taxes and tariffs – though really sure, that too), requiring professionals throughout a number of departments to work in spreadsheets predicting and making selections about shelling out the following six to 12 months of stock. The sheer scale of managing hundreds of provide chains by way of quite a few difficult procedures is a longstanding drawback. It’s additionally meant that the majority organisations might solely expend the assets and time to run S&OP as soon as a month.
“SAS Provide Chain Agent runs constantly to stability demand, provide and operations,” states SAS. “Customers can optimise provide chains in durations of excessive demand, forecast future wants primarily based on utilization patterns and scale back waste and over-ordering. Plus, customers can keep ongoing, close to real-time visibility into provide chain operations, permitting them to constantly faucet their information to make smarter selections, inside or outdoors of a typical planning window.”
Following commercially-driven curiosity
Enterprise customers can work together with the agent by way of a chat expertise that enables them to observe their commercially pushed curiosity and problem-solve each time they’d like.
For example, a person might ask the agent to run a situation (say, a 15% drop in demand) and discover potential outcomes, receiving explanations alongside the best way on how the agent arrived at its selections for transparency and trustworthiness.
“Present pre-packaged brokers are likely to deal with fundamental processes; with Provide Chain Agent, SAS is compressing a really advanced course of, which might ship vital worth,” mentioned Kathy Lange, analysis director at IDC’s AI, information and automation software program follow. “This providing positions SAS to carry its longstanding provide chain data to a brand new technology of agentic AI options.”
First debuted at SAS Innovate 2025, SAS used its SAS Innovate 2026 person and practitioner convention to elucidate how its platform permits customers to create digital twins of consumers’ industrial environments in Epic Video games’ Unreal Engine (UE). These totally digital facility replicas permit clients to simulate situations, making a proving floor for patrons to ask “what if” and work out learn how to act.
As a working instance, in hospital rooms, surgical groups can’t carry out lifesaving operations if their full set of vital medical gadgets (scalpels, clamps and many others.) usually are not sterilised and secure to make use of on sufferers. A significant supplier of medical gadget sterilisation is collaborating with SAS to construct a digital twin of their facility, permitting them to discover and check situations that might stop or sluggish supply of their very important providers and optimise how they run.
This buyer believed that trays of medical instruments had been getting caught in a buffer raise that lined the trays up for cleansing, bottlenecking your complete course of. By rendering their services into digital twins and exploring additional, they found that, in truth, the trays had been, in truth, getting delayed as a result of the buffer raise acted as a central distribution level. By making focused changes, the bottleneck was damaged and manufacturing tempo picked up.
SAS state of enterprise AI
Talking throughout the media briefing session at SAS Innovate 2026, firm CTO Bryan Harris requested the press to think about what “sturdy worth” means within the period of AI. As cloud suppliers have used open supply to commoditise proportion of the infrastructure that builders use to construct enterprise functions, that shift was a serious growth… however a fair larger and extra impactful shift is now underway.
“However AI has basically re-shifted the economics of construct vs. purchase,” mentioned Harris. “As we now work to create sturdy worth, it comes all the way down to governance, agentic AI, digital twins and quantum AI as these forces all come collectively.”
Welcoming Reggie Townsend, VP of ethics, governance and social influence at SAS to the stage, Harris questioned Townsend on the foreign money of belief within the period of AI. Pointing to the work SAS has accomplished with its AI Navigator service, the pair defined how this new service works.
“We wished to handle the concept governance is the trail of most resistance,” mentioned Townsend. “So we wished to make governance irresistible, accessible, intuitive and action-oriented. We need to make it clear that AI governance is a development driver. As an alternative of fears of shadow AI placing the organisation in danger, AI governance empowers folks to push the boundaries of AI inside a structured, clear and safe surroundings.”
SAS AI Navigator gives a unified view of no matter fashions and instruments a group is already utilizing, together with LLMs, AI brokers and open supply or SAS fashions. It helps the journey from experimentation to deployment by retirement, offering a unified view of all ruled belongings, whether or not constructed in-house or bought from third events.
SAS’ Profi: We’re on the level of workinig with AI that ‘acts’ in instruments, methods and workflows.
Jared Peterson, SVP of worldwide engineering, took over the prolonged part of the SAS Innovate media session. Inviting plenty of visitor audio system on stage, together with Marinela Profi in her function as world market technique lead for AI brokers and gen-AI at SAS, Profi talked about brokers are actually taking actions throughout instruments, methods and workflows.
“It’s what I name ‘AI that acts’ right now,” mentioned Profi. “We all know that customers usually are not essentially challenged by fashions themselves, they’re challenged by the whole lot that occurs ‘round’ language [and image, and other] fashions as they begin to apply brokers to some extent of actionable affect on an organisation’s information… and that’s when enterprises must begin to resolve to what diploma that can put brokers in management to make selections and, crucially, the place they put their human-in-the-loop.”
Safeguarding with artificial information
Trying wider into different toolsets, SAS Employee Security permits organisations to handle office dangers utilizing digital twins, artificial information and pc imaginative and prescient.
With this providing, clients use digital twins to create sensible footage for coaching pc imaginative and prescient fashions on high-risk situations. This strategy permits for nearly limitless variation in simulated environments, capturing essential particulars like the form of protecting eyewear, gear color and the way totally different lighting situations can have an effect on an accident.
“Artificial information and pc imaginative and prescient additionally make it potential to mannequin uncommon however believable occasions for which actual footage could not exist, like forklift collisions. Through the use of totally simulated employee personas, organisations can repeatedly check particular sequences of actions with out involving actual staff or exposing any personally identifiable data,” defined SAS, in a technical briefing doc.
As soon as skilled, these fashions will be deployed throughout cameras inside a facility to supply actual‑time alerts, serving to guarantee staff are sporting protecting gear accurately and sustaining a secure surroundings. On a manufacturing facility ground, this would possibly imply verifying correct helmet positioning, or, in medical settings, detecting a slipped masks or glove earlier than a lab or working room is compromised.
SNAP struggles
In administering Supplemental Vitamin Help Program (SNAP) advantages, American states typically wrestle to maintain tempo with evolving laws, heavy caseloads and managing time-consuming handbook tech duties. Now, new federal regulation can instantly advantageous state budgets for exceeding the brink for fee error charges: the proportion of advantages which can be over- or under-awarded due to eligibility miscalculations, outdated case information or undetected fraud. These compounding errors can value states thousands and thousands in federal funding. And, most significantly, households who want very important help is probably not receiving all the advantages they qualify for.
A number of states within the US are utilizing SAS Cost Integrity for Meals Help to confront this drawback and higher serve their constituents.
Posten Deliver used SAS Viya and SAS SingleStore to underpin its high-demand surroundings the place its methods run constantly, 24/7, to assist real-time parcel monitoring, route optimisation and buyer communications..
GLOBAL NOTE: Though this (above) a part of the SAS story is undeniably US-centric, the corporate did use its SAS Innovate 2026 keynote mainstage session as an instance buyer tales at the moment enjoying out world wide, notably in Europe with Posten Bring (a logistics and supply supplier within the Nordics) as a living proof.
“When organisations are left stitching collectively ad-hoc AI frameworks and experiments, they typically fail to realize the aggressive edge they’re in search of once they spend money on AI,” mentioned Manisha Khanna, world market technique lead for utilized AI at SAS. “We’re engineering {industry} accelerators with goal: to unravel outlined, actual {industry} issues in extremely regulated environments. With production-ready brokers and fashions that work on information they have already got, our clients throughout industries can and are reaching extraordinary outcomes.”
SAS’ fashions have been skilled on patterns from a broad dataset contributed by way of consortium by main world monetary establishments. The corporate says its SAS {industry} accelerators are rigorously examined and designed for his or her designated capabilities. Plus, by integrating with an organisation’s current workflows, {industry} vertical specialists can use SAS’ portfolio to increase their analytics and AI capabilities with their current information.
50 years – after which AI
What is maybe most fascinating in regards to the work at the moment being carried out at SAS is the age of the corporate. That is an organisation that’s this yr celebrating its 50-year anniversary – and there are treasured few expertise organisations which have evidenced this degree of longevity. There are even fewer tech companies of any form which can be working with neural networks, trendy governance points, artificial information and all manifestations of agentic AI providers.
The SAS of right now is clearly a major development if we glance again on the organisation’s roots – what began as a government-funded information mission at North Carolina State College to analyse large quantities of agricultural information has pervaded, pivoted and positioned itself as an organization constructed for the info science age that’s now able to straddle the post-cloud influence of agentic AI at each {industry} software level.
