Everyone has had a day at work where they wished they could replicate themselves. Or replicate their top employee. Agentic AI may be the next best thing.
Essentially, AI agents can act autonomously without human oversight, acting within parameters humans have set for them. They “use AI techniques to make decisions, take actions and achieve goals,” said Bradford Loewy, director of product solutions at Bulloch Technologies, during a panel at the 2026 Conexxus Annual Conference in January.
Here’s an example from Clerley Silveira, engineering manager at Invenco by GVR. Today, generative AI can help you plan a vacation by giving you a suggested day-by-day itinerary with possible flights, timelines, activities, places to stay or eat and other logistics. Agentic AI will go a step further. Based on the preferences or rules you give it, it will go ahead and make the bookings for you.
“Once we are at that stage, where you no longer have to take the action yourself, that’s when we’re going to have a true agentic AI. And that is what’s going to be really transformative,” said Silveira.
The Agentic Store of the Future
What does agentic AI mean in a c-store?
Many operators might already be using AI for automation. An example is a forecasting report for inventory that helps optimize what to order and when. A human still has to place the order. Or maybe your AI software can already place orders for you based on reorder triggers, but it can’t take other factors into account and adjust orders accordingly based on shifting dynamics such as weather, events, promotions or supply chain disruptions that are programmed into your system.
“Where agentic AI really starts to play into it is when it makes these tasks become menial. These processes that we count on, the agent can start to take over those for us,” said Michael Munz, marketing operations manager at Petrosoft.
Fuel
On the forecourt and in underground tank storage, Jack Dickinson, director of partner development at Dover Fueling Solutions, said operators are working on both reactive and preventative agents that can start responding to today’s automated systems.
“Think about managing leaks. Now you can look at meters, what’s in the tanks, and look at deliveries coming in and fuel going out. You can do that today and be reactive,” said Dickinson, but the future might hold automated dispatching. Imagine an agent that knows the location and schedules of technicians and can reroute as new priorities evolve.
Even better, imagine an agent that knows everything on a technician’s vehicle and can route according to what parts are likely to be needed at what location.
“Preventative and operational efficiencies can really be automated. You can allow the systems to start making decisions and plug into other aspects of your operations and your systems that they’re not connected to today,” said Dickinson.
Inventory management
“Inventory today is a living system,” said Petrosoft’s Munz. “It’s tied to supply chain realities, seasonal patterns, customer demands, in-store loyalty, manufacturer loyalty, incentives, and increasingly things like shrinkflation and tariffs.”
“AI follows rules. It triggers automations based on learned and trained context, and that’s where we’re going to have to start to do our biggest amount of work in training AI agents. AI will need to understand patterns, predict outcomes, adjust actions accordingly,” Munz said. “So instead of restocking and forecasting just because something hits a number, agentic AI is going to be able to take [everything into consideration]. It might delay or accelerate an order because it detects a coming promotion, a weather event or a shifting customer demand due to shrinkflation or pricing sensitivities.”
Foodservice
Foodservice is one of the most complex operations in retail, said Mike Weber, chief growth officer at Upshop. Many operators are still managing foodservice with spreadsheets, mental math and yesterday’s sales. As margins tighten and labor pressure increases, that approach is becoming unsustainable, he said.
“This method puts the onus on your team to make so many daily decisions in a short amount of time while under a lot of pressure,” Weber said. “The computer should be focusing on those things that can become automated so that your team can focus on what’s more meaningful.” Foodservice involves “thousands and thousands of smaller decisions” that can be taken off your team’s plate so they can focus on elevating performance and customer service.
Weber said that in foodservice operations, AI should be tracing all ingredients from the minute they come in the door, helping operators decide how much food to produce by location and daypart, tracking waste, guiding teams on how to drive markdowns and forecasting inventory needs for the next few days up through a few weeks out.
He said store-level demand forecasting is the most realistic and impactful starting point for AI in foodservice. “There is the opportunity to get up to 98% or 99% accuracy, which will mean you have extremely efficient stores.”
Once agentic AI is turned on and connected to forecasting systems, it will be able to make even better recommendations and optimize foodservice operations. “You can train your agent to focus on your most key metrics so that it is actually anticipating what should change in each of your stores based on the signals that it sees. If the agent keeps seeing a recurring pattern from a supplier, from a forecast or from behaviors in the market, it’s going to start prompting new recommendations,” Weber said.
Mobility Commerce
If you ask Gray Taylor, former executive director at Conexxus, what will define the future of convenience, there is one answer: mobility commerce, and agentic AI is the powering force.
Your car will serve as a connected piece of the convenience ecosystem—you can ask it to stop for gas at your favorite store, create a mobile foodservice order or purchase a pack of cigarettes while it verifies your age. Your car will know your loyalty app and be able to pay for all these products before you arrive. License plate recognition technology will log your car at the pump.
But there’s a lot of work that comes with bringing this to fruition.
“We have, and I can’t emphasize this enough, a freight train coming down the tracks called agentic shopping. It’s being driven by the ability of consumers to find you online, of systems knowing your real-time inventory and what you have in the store and your pricebook,” said Taylor. “And AI will be making pragmatic decisions for the consumer, to go to this store I can make two right-hand turns, but to go to my favorite store I need to make two left turns. AI will start deciding that and learning preferences. So we have a big job ahead of us and its going to be a huge operational challenge.”
The Human Touch
Agentic AI has transformative capabilities, but speakers at Conexxus made one thing clear: We shouldn’t automate just for automation’s sake. “There’s a point where automation becomes counterproductive, especially when it overrides store-level judgment. That’s a human level thing, and we need to value it,” said Petrosoft’s Munz.
Retailers who succeed will be the ones that understand what to automate, what to protect and where humans must remain central in order to protect consumer trust.
“Retailers that win will be those who deploy the tech with clarity, balancing automation with human oversight to build trust by preparing for the risks and evolving guidelines, and recognizing that inventory is just one piece of a much larger AI-powered e-commerce platform,” said Munz. “The power of inventory in agentic AI is what everybody in this room thinks is going to be easy and it’s not. It’s about trust.”
The Next Generation of Security
Without strong security and governance, AI is not innovation, it’s exposure, said J.B. Branch, a big tech accountability advocate focusing on consumer rights and data privacy at Public Citizen.
For a consumer‑facing industry that touches millions of customers every day, using AI as part of core business infrastructure changes the risk profile of the business. “When failures happen, you lose consumer trust,” he warned. “We want to scale AI innovation. It is a positive thing, but you don’t want to erode consumer trust.”
“As we increasingly move towards agentic AI,” he said, “you will have one company with thousands of AI agents and another company with thousands of AI agents that are going to be interacting with one another in cyberspace without much human oversight. And that is concerning if harms or errors happen.”
One of Branch’s strongest cautions was around vendors and data sharing. He urged retailers to understand where their data is stored, which vendors can access it and whether it’s reused to retrain models.
Before deploying AI, Branch said retailers should consider:
- Defining data boundaries (what data AI can and cannot touch)
- Locking down vendor contracts (data use, retention and server location)
- Planning for failure at scale, not just edge cases
- Maintaining human accountability
“You need clear internal rules on what data you’re using, how long you retain it and what it’s used for,” he said.
Putting Yourself Out of Business (in a Good Way)
Having started on the frontlines of his family business, Frank Gleeson, CEO and president of NACS, has a motto: “If you’re not serving the customer, you’re serving someone who is.”
During his Conexxus keynote address on innovation and technology in convenience, he drove home the point that frontline store employees have the hardest job in the organization—and innovation and technology that don’t make their work easier will fail. “Technology is a huge enabler,” he said. “As you think about tech, it’s got to be about making things easier and removing friction.”
The best technology tools, he said, simplify tasks instead of layering on new steps, reduce cognitive load on employees and eliminate unnecessary manual work and redundancies.
In his years as president and CEO of Aramark Northern Europe, he recalled working with a computer-vision self-checkout company in its nascent stages and implementing it in sports stadiums. “It was quite transformative at the time, and we knew that we were looking for something that would make [checking out] easier and faster for our customers, especially for high-volume outlets or outlets where our labor costs were particularly high.”
Importantly, this is not about reducing headcount, he said. Better technology leads to better service, not fewer people. When tools do their job well, employees can focus on what only humans do well.
With successful innovation comes failure—not all ideas are going to work, and they don’t always work quickly. Referencing the self-checkout example, Gleeson told the audience that the deployment took years, not months, to roll out at scale. “And we tried a lot of other new technologies in stadiums that just weren’t successful.”
His advice to retailers when innovating: test small, fail fast and scale only what creates repeat behavior.
In a Q&A with panelists Chris Bambury, president at Bambury Inc.; Lonnie McQuirter, director of operations at 36 Lyn Refuel Station; Varish Goyal, CEO of Vintners Distributors, Au Energy and Loop Neighborhood Markets; and Jigar Patel, vice president of SAASOA USA and CEO at Fastime, the group pointed to examples of where technology delivers real value for operators:
- Mobile ordering and pickup that actually saves time
- Seamless car wash and fuel experiences
- Touchless entry points and payment flows
- Technology that quickly allows customers to customize their orders
- AI‑assisted checkout that speeds transactions
- Automated cleaning and maintenance that frees staff for service
Innovation isn’t always about novelty or being first to market. Instead, it’s about introducing the right technologies that solve the right problem to transform a business.
Twenty-five years ago, the late Steve Sheetz, former president, CEO and chairman of the board at Sheetz, said his vision was to create a version of Sheetz that would put the Sheetz he led out of business.
“How would you put yourself out of business [today] if you were to innovate and grow?” asked Gleeson.
Conexxus is a non-profit organization that sets convenience industry technology standards and defines strategic technology roadmaps to navigate the evolving retail environment, ensuring your operations stay ahead in the competitive landscape. To learn more about Conexxus and how you can get involved, visit conexxus.org or reach out to Linda Toth, executive director of Conexxus, at ltoth@conexxus.org.