When Rick Rigby joined Atlantis Management Group as chief technology officer, he came armed with expertise from nearly two decades at Circle K and a firm belief that better data leads to better decisions.
Rigby began building a data team at the company and creating new dashboards, knowing that Atlantis had valuable information that could help employees improve operations.
“I thought, I’m going to have my ‘Field of Dreams’ moment. If you build it, they will come,” he said while speaking at PDI Connections Live in August 2026.
That moment arrived during a meeting with category managers and company leadership. The CEO of Atlantis was asking highly specific questions to the group. The category managers knew their business, but they didn’t have the exact data in front of them to answer.
“I pulled out my tablet and the dashboards I was working on and showed him the answers he was asking for,” Rigby recalled. “He looked at it and he said to me, ‘Teach them.’”
Over the following weeks, Rigby trained category managers and operations leaders to understand and use the data more effectively.
“They were able to be strategic. Having this data took them from being reactive to predictive. They were able to start figuring out things that they didn’t see before,” he said. “They were seeing patterns that I, as a data guy, didn’t even see. The data was democratized to bring more insights to more people. It was very powerful.”
'Forget the Sexy Stuff'
If you talk to both operators and suppliers across the industry about implementing AI in retail, you’ll start to hear a recurring narrative: “Forget the sexy stuff.”
That doesn’t mean use cases have to be primitive, it just means they should be practical. And practical can still transform your day, your store and your business.
One of the biggest impacts AI can have on a business is that it can do the mundane daily tasks that take up time. And it can often do them faster and more accurately.
Inputting fuel invoices, expenses and other documents is one of operators’ most tedious to-do list items, said Safi Modi, head of growth at Modisoft. “Manually keying in information takes so long. Why not have AI get it done for you?” he said.
AI document readers, including one developed by Modisoft, can scan invoices or receipts, extract the information, log it and reconcile it with the back office.
A task that initially could take up to seven minutes now takes less than one. Added up, that can save a significant amount of time across the week or the month. “You can look at it as staff augmentation,” Modi said.
AI can also flag errors, like if the batch payment matches the initial order but not what was delivered and is now in inventory. “Say you actually received 10,000 gallons of fuel, but you got billed for 12,000 because you initially ordered that amount. AI is tracking how much the invoice was for verses what was actually logged to see if it reconciles or if it’s mismatched.”
Flagged anomalies in sales data can also signal problems or opportunities for improvement. Instead of requiring managers to review reports from every location, AI can identify unusual activity and call attention to it.
“Say last week there was a store that made $2,000 [on a product], but this week it’s only $1,000,” said Modi. “That discrepancy is going to be flagged.”
At Atlantis, Rigby partnered with PDI Technologies to test an AI-powered fuel invoice workflow. Previously, invoices arrived via email, an accounts payable employee manually entered them into the system and another employee routed them for approvals. Multiple people touched the document before payment.
“Now AI can just read a PDF and pull out the seven or eight data points you care about and put it through the approval process,” said Rigby. “That whole process went from three or four people spending days to an automated flow.” Critically, the process still has “humans in the middle for approval.”
“It really moves the needle in terms of reducing labor and helping our teams be more strategic,” Rigby said.
Make Impact With Insights
C-stores collect a vast amount of data, but what do they do with it? Are there employees who have the time and the skill set to analyze the data and build strategy around the numbers?
“It’s all about making data actionable,” said Jeff Hassman, general manager of AI solutions at PDI Technologies.
Today’s AI tools can provide insights, surface opportunities and help determine ways to grow the business.
Rigby described one of his first experiments with AI, asking about a particular store’s sales using a report from his self-checkout vendor.
Prepaid fuel card sales, which were very high across the company overall, were significantly lower at this site than at comparable locations, even though it was a high fuel volume location. “What I realized was that the employees and customers didn’t know they could buy prepaid fuel cards on that self-checkout machine,” he said.
“It was a training opportunity,” Rigby said. “AI was able to give me the answer to a question I didn’t even know I was looking for.”
After additional training, prepaid fuel became one of the store’s top-performing categories.
In other cases, AI can help put strategy behind qualitative information.
Transaction-log data can often tell you what was purchased, but not why, said Kevin Farley, chief customer officer at InStore.ai. InStore.ai’s technology, which analyzes conversations in the store and then uses AI voice analytics to surface operational improvement opportunities, gives retailers insights into what’s really happening in the store in a way that might not have been captured otherwise.
“If customers are asking your cashier for a product five times a day and you don’t carry it, that’s important information,” he explained. Historically, there was no system for capturing those requests, identifying patterns and getting the information back to decision-makers.
“The conversation analysis is finding patterns and turning those patterns into something retailers can act on,” said Farley.
Pick Up the Pace
Innovation is moving at what feels like lightning speed, and “the game has changed,” said Rigby. “So how do you go faster, but maintain the same level of quality [in your operations]?”
In an age where viral social media trends are heavily influencing what’s popular, retailers need to hop on product inventory fast. “If that product has a three-week life cycle and you figure it out in week four, you’ve already missed it,” said Babir Sultan, president and owner of Fav Trip, a three-store chain in the Kansas City, Missouri, area.
Sultan uses AI to turbocharge the category manager function in his stores. Identifying emerging products and in-demand items to stock can require hours spent reviewing trend reports, social media conversations, supplier information and industry news. But AI helped the c-store owner cut that task from four to six hours down to four minutes using Claude.
Each day, Sultan asks the AI agent what the top five items are that the retailer should look at. “It told me to stock items I didn’t carry even a month ago that are now our top sellers,” said Sultan. Rather than forcing store teams to navigate dashboards and reports, AI distills information into practical priorities, he said.
Sultan believes some of the biggest benefits of AI come from helping employees perform everyday tasks more effectively. Anything that simplifies work, reduces frustration or saves time ultimately improves store execution, he said.
“This is especially critical in a retail environment where labor is a challenge and turnover is high,” he said. “Anything you can do to make employees’ jobs easier and faster will make them more engaged.”
Shyam Rao, founder and CEO of Tote, an AI-native POS platform, described how a manager at one retailer the company works with said using the system’s automated features for her end-of-day close out saved her 45 minutes worth of administrative work each day, allowing her to focus on other aspects of the store. “That time saved translates to real ROI,” he said.
AI implementations will be unique to the retailer and the problem they’re trying to solve, and there’s no one prescriptive way to start integrating the technology into your workflow. But at the end of the day, AI tools should solve a problem, provide a new layer of insight, or benefit your employees in a way that drives real impact on the business.
“I think the idea that every employee in your organization is going to be smarter and make better decisions because they’re going to have these AI assistants … that’s where the magic is,” said Hassman.
Beyond the Hype: 5 Rules for Getting AI Right
According to AI consultant and executive strategist Sol Rashidi, retailers should “avoid the shiny toy syndrome” of AI—in other words, spend less time chasing the latest AI trend and more time making sure the technology is solving the right problems. Simply using AI tools doesn’t translate into meaningful transformation, she said during her keynote presentation at McLane Engage 2026. “Don’t ask, ‘How can we use AI?’ Ask, ‘How can we run a better business?’” she said.
These are a few rules she says retailers should follow as they consider AI implementations:
1. START WITH THE PROBLEM, NOT THE TECHNOLOGY
“If my task is to cut a piece of paper, why on earth would I introduce a chainsaw when scissors will do the job just fine?” Rashidi asked.
Before investing in a new tool, it’s important to ask:
- What operational challenge are we trying to solve?
- What’s slowing employees down?
- Where are we losing time or visibility?
2. UNDERSTAND THE DIFFERENCE BETWEEN “USING AI” AND “DOING AI”
A company can deploy hundreds of AI licenses and see employees use them to write emails, summarize documents or generate content faster. “That’s using AI,” Rashidi said.
Those use cases can certainly improve productivity, but she argued they’re often incremental improvements layered onto existing processes.
“Doing AI” is something much bigger, she said. It requires organizations to examine how work gets done, identify where human judgment is essential and determine which tasks can be delegated to AI. It means redesigning workflows and changing how organizations actually operate.
“The distinction makes a massive difference in your approach,” she said.
3. DECIDE WHAT SHOULD STAY HUMAN
Not every decision should be handed over to AI. Organizations need to be “human-led and AI-supported,” Rashidi said.
Retailers need to identify, and differentiate between tasks that are repetitive and rules-based and tasks that require judgment, experience and context. AI should allow employees to focus on the work humans do best while technology handles routine tasks.
4. FIX BROKEN PROCESSES BEFORE AUTOMATING THEM
AI can speed things up … and that means it can also speed up bad processes.
Before adding AI, retailers should examine whether the workflow itself makes sense. Automating an inefficient process often creates a faster version of the same problem.
In many cases, operational redesign should happen before AI implementation, she said.
5. PREPARE THE WORKFORCE
Implementation is not finished when the technology goes live. Employees need genuine AI literacy, including an understanding of what the system does, how it affects their responsibilities and where its output requires human scrutiny.
“Pushing AI into production is not what will determine your success. It’s the adoption that determines your success,” Rashidi said.
