CFVG Embracing AI to Elevate Foodservice

The August CFVG Vision Report examined AI applications in foodservice, covering implementation strategies, data management, and operational transformation.

Embracing AI to Elevate Foodservice CFVG VISION REPORT NO. 3 SEPTEMBER 2025

CFVG MEMBERS

Ryan Blevins Weigel’s

Joe Brumfield La Lomita Inc

Richard Cashion Curby’s Express Market

Jon Cox McLane Company, Inc.

Heather Davis Parker’s Kitchen

Brandon Frampton Loop & Poppy

Stephanie Galentine Lassus Bros. Oil, Inc.

FEATURED SPEAKERS

FACILITATOR

Kris Klinger Boston University

Jac Moskalik Global Partners, LP

Jasmine Struble Yesway

Derek Thurston Cliffs Local Market

Mike Weber Upshop

Bonnie Zaring RaceTrac Inc

VISION GROUP NETWORK CO-FOUNDERS

Mike Weber CMO, Upshop

Steve Brask Director of Business Intelligence, Upshop

Richard Poye COO, Food Trends Think Tank

Myra Kressner Kressner Strategy Group

Eva Strasburger StrasGlobal/Compliance Safe

Roy Strasburger StrasGlobal/Compliance Safe

We are Convenience Foodservice Vision Group (CFVG), a group of invited leaders from across the foodservice sector who have volunteered our time to help our fellow culinary, marketing, operations, technology, sensory, safety, and select solution providers. The only reason we gather is to discuss, debate and share our experiences and ideas. Each of us is offering our personal opinions. We are not looking for “group think.” We make our conversations available to everyone in the industry through CFVG Vision Reports. These reports will help you better understand current challenges, solutions, and opportunities while giving you access to different opinions and perspectives. Convenience Foodservice Vision Group is part of the Vision Group Network, whose mission is to gather the best minds in the industry, put them in a virtual room, and let the ideas and opinions develop. This CFVG Vision Report is comprised of multiple parts: CFVG Views is a summary of the conversation with additional resources. The full meeting transcript and the presentation recording and slides are online and searchable by topics. We include the full transcript so that you can be in the room with us, rather than only having access to selected quotes and paraphrasing.

The main topics in this CFVG Vision Report:

Demystifying AI: Practical Paths for Foodservice and Retail Turning Hesitation into Action • Practical AI Applications and Use Cases

• Preventing Waste with AI • Data Management and Sourcing • Employee Adoption and Change Management • Implementation Logistics and Future Outlook

www.vgnsharing.com

CFVG thanks our Ally Supporters

Founded in 1894, McLane Company Inc. is one of the largest distributors in America, serving convenience stores, mass merchants, and chain restaurants. As an industry- leading partner to the biggest retail and restaurant businesses, McLane buys, sells, delivers, and serves the world’s most beloved brands. With headquarters in Temple, Texas, McLane has more than 80 distribution centers across the country, employs more than 25,000 teammates, and delivers to nearly every zip code in the U.S. McLane is a wholly owned subsidiary of Berkshire Hathaway, Inc. McLane Fresh, McLane’s retail foodservice program, offers unmatched value to convenience store operators by forming strategic partnerships based on tailored insights. The McLane Fresh team provides customized recommendations and seamless execution. Leveraging extensive data, McLane supplies programs and products that drive profitability, focusing on efficiency, transparency, and predictability. Their easy-to-execute packages minimize the need for skilled labor, boost sales with merchandising strategies, and include the right equipment bundles. McLane offers quality products at competitive prices, adhering to top food safety standards, along with fast, fresh and cost-effective deliveries. CONTACT: Jon Cox Jonathan.Cox@mclaneco.com McLaneCo.com

Transform your operation with Upshop’s foodservice operating system - the first platform that seamlessly connects all store workflows with an AI-powered replenishment system. From inventory and ordering to food production and waste management, unlock unprecedented efficiency and insights. Imagine the possibilities when your entire foodservice operation speaks the same digital language, driving profitability and growth across every location. CONTACT: Mike Weber mike.weber@upshop.com upshop.com

www.vgnsharing.com

CFVG VIEWS

This meeting of the Convenience Foodservice Vision Group (CFVG) focused on clarifying the practical applications and potential uses of AI in foodservice and convenience retail. A CONVENIENCE FOODSERVICE VISION GROUP DISCUSSION

The CFVG quarterly virtual meeting, held on August 20, 2025, was facilitated by Richard Poye , chief oper- ating officer at Food Trends Think Tank. The meeting featured guest speakers from Uptop Mike Weber , CMO, and Steve Brask , director of business intelligence, with their presentation “Winning with AI,” which discussed how businesses can approach AI transparently and understand its capabilities and journey to unlock new possibilities. Weber, a CFVG member and Ally Supporter, invited Brask as his guest and co-presenter as he is known within his organization as the go-to resource for numbers and analytics. Weber and Brask proposed approaching AI in a practical, brand-agnostic way to help organizations at varying stages of adoption to identify opportunities for growth.

www.vgnsharing.com

Demystifying AI: Practical Paths for Foodservice and Retail From the outset, Weber positioned his presentation as a chance to strip away the mystery around AI. As he put it, “how you approach AI should not be a secret. It should not be kept inside of a black box of […] a technology partner or a consulting firm.” His goal was to help attendees grasp what AI can do today and, more importantly, how each business can shape its own AI journey. At Upshop, he acknowledged, the team continues to explore the “realms of possibilities” for AI, guided by a belief in the “art of the possible.”

Weber underscored the stakes with a clear message: “AI is rewriting the rules of pretty much everything,” a shift that will profoundly reshape foodservice and retail. To ground that vision, he outlined three ques- tions at the heart of the discussion: “What should AI do?” “What does that really look like?” and “Where do I start?” He illustrated the daily challenges of foodser- vice, from food safety and disconnected tech to task overload and costly blind spots, using the familiar rhythm of prep and production, where staff are left wrestling with constant “mental math” about how much to make or how much more they could sell.

Ideally, he explained, AI should take on those bur- dens: forecasting demand by location and day, guiding production, optimizing markdowns, man- aging waste, and even enhancing traceability for high-risk ingredients, nutrition facts, and allergens. Weber walked through AI’s history, from 1950s com- putation to 1980s expert systems, 1990s statistics, 2010s machine learning, and today’s explosion of generative AI. The turning point, he noted, is shifting from asking “Did I reduce shrink today?” to “Show me how to reduce shrink and what are my options?” The ultimate aim, he said, is to “optimize thousands of these big and small decisions every day to drive performance and to do this in a relatively small amount of time.” From there, Weber and Brask introduced a four- step framework for embedding AI: introduce, train, partner, and autopilot. The big ‘aha’ has obviously come in the past couple years with GenAI. And instead of using rules or relationships or patterns, we can now simply say, given an outcome, discover my inputs. Mike Weber, CMO, Upshop

AI is rewriting the rules of pretty much everything and it certainly has and will continue to have massive impact on foodservice and the business, your businesses. Mike Weber , CMO, Upshop

www.vgnsharing.com

At some point, there is the ability for AI to start drawing conclusions and to offer recommendations, or at least at minimum, to provide different scenarios based on what’s happening in your business. Mike Weber, CMO, Upshop

Introduce AI to Your Data: The jour- ney begins with AI-powered demand forecasting. Brask explained how this model blends historical sales, promo- tions, price shifts, and even external factors like weather or holidays. Importantly, he emphasized that “even with poor quality data, the nature of machine learning and AI in general can bridge the gaps that exist,” mak- ing forecast accuracy goals of 95% or more attainable. Train the System: Here, organiza- tions identify the right metrics and build what Weber called their own “approach to Moneyball.” By tracking plan engagement, forecast compli- ance, and menu compliance, busi- nesses can shift behavior in stores while AI fine-tunes production sched - ules around shelf life, sales patterns, and timing. Partner with AI : At this stage, teams become digitally empowered. Task management evolves from rigid

schedules to dynamic, AI-prioritized workflows. With large language mod - els (LLMs) layered over enterprise data, employees can “ask questions in whatever language [they] feel com- fortable with and then get actionable insights out of [their] own data,” Brask said. Autopilot : Finally, AI begins to draw conclusions and model out different scenarios. Specialized AI “agents” for supply, limited time offers, or waste can proactively suggest strategies or flag issues, even for events with no precedent, like a new store opening or product launch. Weber wrapped up by inviting ques- tions, stressing that these four steps form the foundation of any roadmap for embedding AI. The conversation he noted, is about practical possibili- ties, meeting businesses where they are today while preparing them for a future in which AI is central to every decision.

www.vgnsharing.com

Turning Hesitation into Action

After the presentation, CFVG members dove into a discussion about how AI is taking shape in the retail world, especially for small and mid-sized businesses. The conversation touched on everything from retailers’ hesitations and

worry among retailers that they “keep thinking they need to be Google” regard- ing their data quality to even contemplate AI adoption. Weber emphasized how AI can help retailers manage complexity

the many ways AI can be applied, to the hurdles of handling data and helping employees adapt. Members also looked ahead at what AI might mean for the future of retail. A common thread throughout was the importance

across multiple stores, even when those stores use different point-of-sale (POS) systems and produce inconsistent data. He explained that AI models are designed to shape reliable forecasts from fragmented infor- mation, allowing operators to quickly explore

of breaking down the complexity of AI and showing how it can be practical, approachable, and genuinely useful across all kinds of retail operations.

scenarios such as the potential impact of a limited time offer or the performance of newly acquired stores. These models, he noted, can learn and adapt on limited data, offering valuable insights in real time.

A central point of concern for many attendees revolved around the perceived high barrier to entry for AI, particularly for businesses that might not have the extensive resources or perfectly curated data of larger corporations. Poye articulated this sentiment, stating, “I think about some of the smaller retailers that are out there that might read this report, and they’d be like, ‘I’m over- whelmed by this.’” This immediate apprehension led Roy Strasburger , CEO of StrasGlobal, president of Compliance

Brask expanded a concept Upshop internally calls “global training,” which allows machine learning models to generate accurate demand forecasts even when data is incomplete, incon- sistent, or limited to certain locations. He noted that while different names are used in the field, Upshop adopts “global training” because it is the clear- est and most approachable description. For example, menu items that rotate in and out or are offered only at select stores can still be forecasted across an entire network. Drawing from his experience in mul- tiple implementations, Brask noted that smaller, tightly knit teams often achieve stronger outcomes because they minimize noise in the adoption process.

Safe, and Vision Group Network co-founder, to directly seek reassurance from Weber, expressing concern about the potential overwhelm small and medium retailers might face when adopting AI. He highlighted the widespread

www.vgnsharing.com

made-to-order pressure in the kitchen, making ser- vice faster and easing the workload for employees. While she acknowledged their existing data isn’t “beautiful” or easy for third parties to interpret, it has been sufficient to guide these adjustments. The priority, she noted, has been creating a manageable workflow for staff since the size of the menu can be overwhelming. After working a shift herself, Gal- entine realized firsthand how antiquated some of their systems are, from back-of-house operations to something as basic as the label maker. Listening to Weber’s presentation, she said it was eye-opening, prompting her to consider whether to bypass incremental fixes and instead “leapfrog” directly into AI solutions to better manage pricing, waste, and production. “We do a lot of great things, and we have great food,” she said, “but the way we do it is antiquated.” Practical AI Applications and Use Cases The discussion also covered tangible examples of how AI can be applied to solve real-world retail challenges, spanning inventory, waste, and strategic planning. Poye asked Jon Cox , vice president of retail food- service at McLane, if the company is engaging any

He emphasized that retailers do not need to match the scale of companies like Google to see results; AI can deliver value even from small datasets, particu- larly when global training is applied. Strasburger then raised a common concern among retailers: whether incomplete or low-quality data can support an effective AI strategy. Brask reassured him that it can, pointing to a recent example where his team helped a partner forecast demand for items last sold four years earlier, during the COVID-19 era. Despite relying on outdated data, the model, through global training achieved 87% accuracy. He added that as teams adopt AI tools, they generate more reliable, up-to-date data, which in turn strengthens the model over time. Later in the discussion, when asked how her team was managing a complex menu alongside grab- and-go and touchscreen ordering, Stephanie Galentine , COO of Lassus Bros. Oil, admitted they are not yet using AI. “I still feel like I’m 10 years behind,” she said candidly. However, she said the possibilities have her rethinking her approach. She explained that her store has streamlined its warmer, focusing only on items that sell and applying deeper discounts to move product quickly. This reduces

What global training allows us to do is take that limited dataset and generate a demand forecast for that item across all of our locations, regardless of how much data is available. Steve Brask , Director of Business Intelligence, Upshop

www.vgnsharing.com

retailers from an AI point of view. Cox, while noting they aren’t directly engaging retailers with AI services yet, they are receiving questions from them. He described AI’s potential in inventory management through order predictability, envisioning AI assisting store teams by automatically generating forecasts based on recurring events or patterns and used a real-life example of a bicycle convention causing a 10% sales increase annually in a store location. “At some point, AI is going to help us write the forecast so that [the] store team doesn’t have to remember [the convention],” he said, explaining that the goal in this case would be to use AI for better prediction of product needs and timing, even proactively informing suppliers of demand eight weeks in advance to ensure stock. Kris Klinger , vice president of auxiliary services at Boston University, described BU’s integrated AI system for waste tracking and demand forecasting. Their system captures both pre- and post-consumer waste using Metafoodx camera technology, allowing them to monitor discarded food and adjust production accordingly. This closed-loop approach links forecasting, production, and waste reduction, enabling the university to predict demand for individual menu items across multiple dining stations with a high degree of accuracy. Klinger emphasized that the AI system is fully integrated into operational workflows, replacing manual methods like “dumpster dives” with a cleaner, more efficient process that also supports cost savings and sustainability goals. Roy Strasburger asked whether the waste tracking focused solely on production or included con- sumer behavior. Klinger clarified that it captures both tracking unused inputs and what students discard, providing insights into consumption patterns and opportunities to optimize operations. Eva Strasburger , president of StrasGlobal, CEO of Compliance Safe, and Vision Group Network co-founder, inquired about food forecasting, prompting Klinger to explain that BU evolved from rough estimates based on meal plan numbers and historical attendance to a granular AI-driven system. The AI now predicts demand for specific items at specific stations, tying forecasts directly to waste tracking to create a fully informed, adaptive system. Looking ahead, Klinger noted that BU aims to expand AI’s role to automate ordering and inventory management, demonstrating early success in applying predictive models to optimize operations across the university’s dining services.

www.vgnsharing.com

Preventing Waste with AI Addressing the “lack of specificity” in c-store data, where many items are sold under a single price look up (PLU), Bonnie Zaring , executive director of food programs and offers at RaceTrac, posed a critical challenge: “We sell a lot of things as a one. And some- times being able to take that and make it actionable when you don’t actually know what was in the cup, or you don’t actually know when it was all rung up the same item.” Brask provided a solution, explaining how waste data can be used to “back into” the vari- eties for items like roller grill products or doughnuts sold under a single PLU, even though customers are choosing from multiple flavors or varieties. By lever - AI is a frequent topic within Vision Groups. Here are Vision Reports from 2025 that relate to AI, especially in foodservice. GCVG Vision Report Practical AI Applications in Convenience Retailing , August 2025 CxVG Owning the Numbers: Data-Driven Decisions from Strategy to Store , May 2025 GCVG Do You Want Chips With That? , April 2025 CLVG Lead the Future: Transformative AI Strategies , March 2025

aging waste data, Brask noted, AI can back into those varieties, predicting, for example, that while a store may sell 400 doughnuts in total, waste capture data can be used to break that down into forecasts for specific flavors. He added that this approach is especially valuable for c-store part- ners managing items like roller grills, doughnut cases, or steam tables, where products are often sold by weight under one code. Poye noted that this method could also provide insight into shrink. By comparing waste capture with sales data, retailers could identify where losses occur, whether on pizza, roller grill items, or other categories. In this way, AI not only supports more accurate forecasting but also highlights opportuni- ties to reduce shrink and improve efficiency.

Weber acknowledged that while there isn’t a direct answer yet, the future holds promise for AI to leverage imagery from security cameras to answer broader business questions beyond simple production forecasts. He envisioned AI looking at the “total store” to address these “much bigger questions.” The discussion also touched on AI’s potential for portion sizing, as Roy asked Weber and Brask if AI could help understand the gap between what consumers think they want and what they actually consume, thereby reducing post-production food waste. Brask shared an experiment conducted with a partner on pizza production, where whole pizzas were cooked and sold by the slice. The part- ner wanted to test whether shifting from batching

www.vgnsharing.com

Employee Adoption and Change Management

whole pizzas to half pizzas and adjusting packaging from two-slice and single-slice containers would impact demand. By piloting this approach across a set of stores, they were able to analyze whether smaller package sizes encouraged more sales of different varieties, created a “halo effect” on other items in the assortment, or changed overall pur- chasing behavior. These kinds of experiments, even on a small scale, demonstrate how demand forecasting data can be leveraged to uncover new opportunities and guide operational decisions. Data Management and Sourcing Data, the foundational requirement for AI, was a consistent topic of discussion, with participants seeking guidance on how to manage and source it effectively. Robert Hampton , VGN’s director of strategic growth and initiatives, specifically asked for advice on overcoming the fear that “my data’s not clean enough.” Brask advised retailers to focus on gathering at least two years of histor- ical sales data from POS transaction logs (TLOGs) as the primary foundation for AI, even if the data is incomplete. He clarified that AI can effectively bridge many data gaps, detect SKU rationalization changes, and learn from previous item demands to generate updated forecasts.

An important distinction made by Brask was that while sales data is the core, external data like weather and promotions can often be layered in by AI providers, reducing the burden on the customer’s internal systems. Weber reinforced this, noting that supplemental data streams, such as weather patterns or local events by ZIP code, are now “much more accessible” and affordable, allowing AI to “build the picture of my business without me having the data.” This accessibility makes robust AI implementation feasible for medium-sized businesses without requiring extensive internal data infrastructure beyond basic sales history.

A critical and human-centric aspect of AI integra- tion is its impact on the workforce and the inevita- ble resistance to change. Eva probed this directly, asking about employee feedback on AI systems: if they find it intuitive, easier or if it makes their jobs more interesting. Weber responded positively, not- ing that AI-driven systems are often more intuitive and easier to use, akin to opening an app without formal training. He observed that guided AI tasks reduce “attack[s]” or surprises for associates, mak- ing their jobs feel easier and leading to increased

www.vgnsharing.com

The solution proposed to engage employees and mitigate resistance was through training and the development of prompting skills. As AI works with user-provided direction, known as prompts, honing the perfect prompts is key to getting accurate and desirable AI solutions. Eva noted that “there are some great courses, which are not very long on how to become prompters,” recognizing it as a burgeon- ing career path. Poye suggested that training long- term employees, who possess invaluable customer knowledge, in AI prompting could foster better engagement and more insightful application of the technology. Brask added a unique perspective: AI models themselves learn from user input, including from skeptics, adapting their responses to be more meaningful, and helping to win over reluctant users. Implementation Logistics and Future Outlook The conversation also explored the practical aspects of AI implementation, staffing, and the broader future trajectory of AI in retail. Brandon Frampton , director of fresh food at Loop & Poppy (Loop Neighborhood Market), posed several direct questions concerning the differences between AI models, the time required for implementation across

satisfaction, especially when they see tangible results like reduced waste. Brask, drawing on his retail management background, emphasized that AI significantly reduces “task overload,” freeing employees to focus more on customer interaction rather than being overwhelmed by competing oper- ational priorities. However, the reality of employee resistance was candidly acknowledged. Galentine shared her personal struggles with long-term employees con- fessing, “Some of my people have been doing it for 15 years the very exact same way. So, for me to say, ‘That doesn’t work, that’s never worked, it’s always been wrong,’ it’s offensive.” She highlighted the difficulty, despite her team’s care, and the neces - sity of slowing down implementation processes to ensure buy-in. Myra Kressner , president of Kressner Strategy Group and Vision Group Net-

work co-founder, questioned Galentine on how she addresses such pushback. Galentine said she was “there now,” and described the challenges of implementing a new product process, emphasiz- ing that getting long-tenured employees to adopt changes has been the most difficult part. While a few trailblazers embraced the new approach, many resisted due to habits formed over 15 years. She highlighted that the process required patience, frequent conversations, and adjustments to ensure buy-in, ultimately slowing the rollout but fostering employee engagement and collaboration. Klinger echoed these sentiments, pointing to blind- ers among managers who “don’t know what they don’t know,” and hourly employees who fear job loss or lack understanding of the technology. He underscored that overcoming this resistance “takes a bit longer sometimes.”

www.vgnsharing.com

unaware of the questions they could or should be asking. This familiarity can make it difficult to gain support for new technologies. The second challenge is that hourly staff may also be hesitant, either because they don’t understand how AI works or because they fear it could replace their jobs. Klinger emphasized that educating teams on the benefits of AI is critical, but adoption often takes longer than anticipated, particularly outside highly engaged groups. Zaring added that another challenge arises from the divide between business experts and data or analytics specialists. While business experts understand operational complexity, data teams excel at modeling and interpreting infor- mation. Misalignment can occur when AI tools fail to reflect real-world needs, leaving business users unsure how to apply outputs or feeling that the models miss essential details. Zaring stressed the importance of identifying critical data points and creating tools that meet non-negotiable requirements to produce actionable results. Poye concluded that each business faces its own silos, and building a frame- work that harmonizes AI adoption across teams, ensuring both sides are aligned at each stage, can help mitigate tension and maximize the value of AI implementation. A challenge that we experience is there are business experts who understand the complexity of the business they’re in and then there are data and analytic experts who are very well-versed in synthesizing and understanding data... And I find the two sometimes are not in sync, or we’re working more in silos. Bonnie Zaring , Executive Director, Food Programs, and Offers, RaceTrac

150 locations (100 with foodservice), and the availability of skilled personnel. Brask addressed the model question, stating that various AI models (like OpenAI, Grok, Claude) continue to leapfrog one another, with OpenAI currently “playing catch-up.” Regarding implementation time, Brask estimated that from receiving data, Upshop could go live with demand forecasting within approximately 14 days, with this period primarily used for evaluating model accuracy. However, he clarified that while the initial machine learning model can be built in “weeks,” refining all inputs such as promotions, pricing, and sales data to optimal per - formance typically takes “six to nine months.” On the staffing front, Brask noted the availability of “lots of data scientists who are more than capable of building a demand forecast model.” However, Weber offered an alternative, suggesting that individuals skilled in prompting the different AI tools can add significant value, potentially more cost-effectively than traditional data scientists, especially for tasks like user interface (UI) workflow changes. Poye connected this to Galentine’s point about experienced employees, sug- gesting that training them in prompting would foster better engagement and leverage their insights into customer behavior and events. He highlighted AI’s potential to help manage competitive intrusion by adjusting forecasts, pricing, or promotions. He concluded that prompting is a wise investment as it’s not complex and allows employees to engage with relevant business questions. He then asked participants about the challenges they face within their organi- zations regarding AI adoption, both personally and professionally. Klinger highlighted two main challenges. The first is overcoming resistance from both managers and frontline staff who are accustomed to long-standing processes. He noted that experienced personnel often operate on “auto-pilot,”

www.vgnsharing.com

Regional differences in AI adoption were also men - tioned. Frampton, based in the Bay Area, observed that his region is “far advanced from the rest of the country” with “everybody here is using AI throughout the day for different reasons.” He expressed a strong conviction that AI “is going to change this industry more than anything ever has” and committed to going “all in” both personally and professionally. Klinger corroborated high adoption in the Boston area, influenced by institutions like MIT and Harvard. He described how Boston University has “dove in both feet first” with AI task forces and initiatives, building a data science facility to apply research. Poye recalled an MIT course advising to treat AI (like ChatGPT) like an intern: providing solid advice, spe- cific details, context, and objectives to help it “think.” He also humorously mentioned the advice My take on it is it’s going to change this industry more than anything ever has. We’ve had a lot of different things come and go, but we feel that it’s going to change every aspect, not just food, but how we train, how we track things and everything.

to “treat the thing well.” Klinger noted that he finds it interesting that younger generations, with different communication styles, might struggle with proper prompting compared to older generations used to clear sentences. Poye expanded on this, suggesting that regional differences in language and questioning could impact AI interaction in multi-state operations. Finally, Weber summarized the overarching future vision for AI in retail: “The differentiator will come in when you have a single data lake , and then you have AI connected across all the actual workflows.” He emphasized that while individual AI initiatives are manageable, the real challenge for small to medium businesses is to achieve a “one data language” across all disparate systems, enabling them to communi- cate and unlock AI’s full transformative potential.

Final Reflection AI in foodservice and retail is less a distant innovation and more a practical tool that can transform daily operations. Concerns about data quality or resource limitations can be addressed through approaches like “global training,” which generates accurate forecasts even from incomplete data. The human element is equally crucial. Long-tenured employees may resist change, requiring patience, communication, and training, particularly in AI prompting to empower staff and align insights with business expertise. Looking ahead, the potential of AI lies in unifying disparate data streams into a single, connected framework. When systems can communicate across workflows, AI becomes not just a technology, but a strategic enabler, reshaping efficiency, decision-making, and competitiveness in foodservice and retail.

A data lake is a centralized repository that stores vast amounts of raw data in its native format. Unlike a data warehouse, which typically stores structured and formatted data, a data lake can handle structured, semi-structured and unstructured data. This allows organizations to store data as it is, without needing to structure it first, making it flexible for various analytical and machine learning purposes.

Brandon Frampton , Director of Fresh Food, Loop & Poppy (Loop Neighborhood Market)

www.vgnsharing.com

NOTABLE AND QUOTABLE

“ Folks are so used to doing the things that they’re doing, the way that they’re doing, using the systems that they were using for ten, twenty, or thirty years, it’s difficult sometimes to get the adoption or the support that you need, particularly from the operations folks. And then level that down to the hourly folks that are supporting it. And they are afraid for a couple of reasons. One is they don’t understand the technology and what it does and how it works, but two is that it’s going to take their jobs away. Kris Klinger , Vice President of Auxiliary Services, Boston University “ We’re going to pick a partner to go all in with. And then personally, I’m going all in as well. I think it’s going to change everybody’s life. Brandon Frampton , Director of Fresh Food, Loop & Poppy “ How you approach AI should not be a secret. It should not be kept inside of a black box of a […] technology partner or a consulting firm. Mike Weber , CMO, Upshop

“ You’re building a very differentiating foodservice strategy, so that will drive what metrics matter to you, but this is where you need to then train the AI to help you identify scenarios for it to optimize those measures. Mike Weber , CMO, Upshop

“

I’ve had a couple of trailblazers that have been willing to be uncomfortable and actually go do it the way we’ve designed, and so we’ve got some leaders that are giving us support. But I would argue that’s the most difficult part is getting everybody on board and using the process as designed, being willing to adjust. And I was very frank, I’m like, ‘If you guys can fix it or make it better than this, I’m all in. Tell us how to do it.’ Stephanie Galentine , COO, Lassus Bros. Oil

“ As long as you have a reasonable amount of historical data, even if that data is incomplete, AI can close a lot of the gaps. Steve Brask , Director of Business Intelligence, Upshop

“ What we think over time is that AI is going to help us better predict what product we need and when we need it. Even to the point of being able to go back to some suppliers and say, ‘You can expect a lift eight weeks from now based off of all the data we have, so if you don’t have yellow paper to wrap the Peanut

“ It might be interesting to go and identify best practices at each one of these stages when you build out those quadrants, what’s needed before you move to the next one so that everybody moves along and understands some of the challenges. Richard Poye , Chief Operating Officer, Food Trends Think Tank and CFVG Facilitator

M&Ms, you should go get it.’ Jon Cox , Vice President of Retail Foodservice, McLane

www.vgnsharing.com

Embracing AI to Elevate Foodservice CFVG VISION REPORT NO. 3 SEPTEMBER 2025

TRANSCRIPT AND PRESENTATIONS

In The Room Transcript The full meeting transcript is online and can be searched by keyword so that you can be “in the room” with us, rather than only having access to selected quotes and paraphrasing

Meeting Presentations and Demonstration Videos

www.vgnsharing.com

VISION GROUP NETWORK CO-FOUNDERS Myra Kressner CEO, Kressner Strategy Group

Eva Strasburger President, StrasGlobal; CEO, Compliance Safe

Roy Strasburger CEO, StrasGlobal; President, Compliance Safe

Richard Poye CFVG Facilitator/COO, Food Trends Think Tank

MEMBERS

Ryan Blevins Dir of Food and Beverage Innovation, Weigel’s

Joe Brumfield Sr Category Mgr – Beverages, La Lomita Inc

Richard Cashion COO, Curby’s Express Market

Stephanie Galentine COO, Lassus Bros. Oil, Inc

Jon Cox President, Retail Foodservice, McLane Company, Inc.

Heather Davis Senior Director of Food Service, Parker’s Kitchen

Brandon Frampton Director of Fresh Food, Loop & Poppy

Kris Klinger VP, Auxiliary Services, Boston University

Jasmine Struble Senior Category Manager, Yesway

Derek Thurston Director of Foodservice, Cliffs Local Market

Mike Weber CMO, Upshop

Bonnie Zaring Executive Director, Food Programs and Offers, RaceTrac Inc

Jac Moskalik President of Food, Innovation, and Strategy, Global Partners, LP

www.vgnsharing.com

The Convenience Foodservice Vision Group (CFVG) brings together a diverse range of industry leaders across culinary, marketing, operations, technology, sensory, safety, and select solution providers. The group addresses menu development, product innovation, operational efficiency, foodservice profitability, technology integration, customer experience, consumer trends, health and wellness, FSQA, supply chain management and more. The group is committed to sharing its views and perspectives to advance convenience retailing. CFVG operates under the Vision Group Network, which gathers the collective knowledge and ideas of its members to create a legacy of sharing within the retail community. For more information and to sign up for future Vision Reports, visit our website: vgnsharing.com/vision-report-library

A part of Vision Group Network

For more information about Vision Group Network email us:

Myra Kressner myra.kressner@vgnsharing.com

Eva Strasburger eva.strasburger@vgnsharing.com

Roy Strasburger roy.strasburger@vgnsharing.com

www.vgnsharing.com

VGN Vision Report Library

The VGN Vision Reports will not only provide you with insight into the issues that leading retailers are concerned about but will give you a vantage point to “listen” to their conversations. Learn about the specific solutions or programs that VGN members are utilizing – and the ones that didn’t work. VGN wants to provide real world solutions for real world issues.

vgnsharing.com/vision-report-library

www.vgnsharing.com

We Value Your Feedback VISION REPORTS SURVEY Please take a moment to share your thoughts with us on our Vision Reports.

SUBSCRIBE Subscribe to receive notifications for all new Vision Reports

FOLLOW Follow us on LinkedIn

Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16 Page 17 Page 18 Page 19 Page 20

vgnsharing.com