By Rick Hynum

Instead of fighting to the death in 2001: A Space Odyssey, what if astronaut Dave Bowman and HAL 9000 had partnered up on a high-tech interstellar pizzeria? The spaceship’s AI-powered supercomputer could have handled virtually all of the operations, from taking orders to marketing the restaurant, managing the customer database and loyalty program, even dispatching delivery pods to mining colonies on nearby moons. That would have kept HAL 9000 too busy to hatch murderous plots, while Bowman developed the recipes and bossed the pizza making bots around.

Conflict resolved—and nobody dies, gets unplugged or evolves into a gigantic fetus. Why didn’t Stanley Kubrick and Arthur C. Clarke think of that?

It’s a silly idea, but not all that far-fetched in 2024. AI can’t run a pizzeria entirely on its own—not yet, anyway. What it can do, though, is remarkable. “AI can be thought of by restaurants as a digital employee,” says Vincent Chaglasyan, the COO of RestoGPT AI, a Los Angeles-based marketing and sales platform for small- and medium-size restaurants. And it could be the most efficient employee you’ve ever had—the kind that works smarter and harder.

“The tasks AI can now perform well independently are ones that involve automation, with the task parameters set by the restaurant, like accepting orders as soon as they arrive or finding a driver for delivery 20 minutes after an order is accepted,” Chaglasyan says. “Yet we’re now also entering times of autonomy, where, with time given to learn behavior, a restaurant’s AI employee could produce independent decision making, like generating a promo offer to be sent to a customer who last ordered two weeks ago.”

In fact, there is so much AI can do for a restaurant, many independent operators might better ask what it can’t do. And, more to the point, “When can I start using it?”

In addition to being a fully functioning pizzeria, Flintridge Pizza Kitchen serves as a test laboratory for AI-driven technology developed by RestoGPT AI. Restogpt ai
AI in Action

If you’re thinking AI is too complex a tool for an independent pizzeria, Chaglasyan will point you to Flintridge Pizza Kitchen in La Cañada, California. It’s owned and operated by RestoGPT AI itself. And, he says, “Yes, it’s a fully functioning pizzeria and a go-to gem for locals in the area.”

Chaglasyan and his team of AI wizards started Flintridge Pizza Kitchen to make sure RestoGPT’s advanced online ordering and delivery software is up to snuff for other smaller operators who want to adopt it. “Anything released by RestoGPT AI is first fully tested and actively used in our pizza shop, from integrations to automation,” Chaglasyan says.

Flintridge Pizza Kitchen is largely a carryout and delivery restaurant, but it has a few dine-in tables, too. And whatever RestoGPT AI can do at Flintridge, it can do for other independent pizzerias, Chaglasyan says. “We walk the walk,” he notes.

With RestoGPT’s PizzaBox ordering platform, Chaglasyan says indie pizzerias just need to provide a link to their menu, and they can start accepting AI-driven online orders within 24 hours via their own fully branded online storefront, also built automatically by RestoGPT AI. Using its high-powered software, he notes, RestoGPT automatically “converts sales data into a structured customer database and activates marketing campaigns to engage and improve customer retention plus the order frequency.”

With a system like RestoGPT’s, this AI “employee” can manage off-premise orders across all channels—direct storefront, third-party apps, phone answering, etc. Additionally, he says, “It dispatches [your own or third-party] drivers and builds customer profiles from these orders for loyalty marketing, resulting in newly generated orders.”

“AI can be leveraged by smaller restaurants and chains. Many of the tools they use today are introducing AI features to help them make better, faster and more accurate decisions.”

Sue Pittacora, Wavicle Data Solutions
AI Use Cases

Whatever data you’ve collected for your store’s operation, AI can put it to optimal use. And the more you have, the more ways it can be used, says Sue Pittacora, chief strategy officer for Wavicle Data Solutions in Oak Brook, Illinois, which counts McDonald’s and other global chains as clients.

Pittacora offers a litany of AI use cases already being explored by major pizza brands: predictive sales forecasting and demand planning based on everything from historical sales data to weather patterns; tracking inventory; predicting staffing and scheduling needs, right down to individual shifts; chatbots and automated order taking; and personalizing digital marketing, including campaigns tailored to customers’ individual tastes and behaviors.

But there’s a catch, of course. “The restaurant, chain or parent company must have the right data, AI-ready data architecture and strategy in place to make these possibilities a reality,” Pittacora adds. “But AI can also be leveraged by smaller restaurants and chains. Many of the tools they use today are introducing AI features to help them make better, faster and more accurate decisions. Additionally, tools like ChatGPT can be used at little to no cost to help with everything from ad copy ideas for promotions and even crafting responses to reviews.”

Dynamic pricing is another option if you think your customers are ready for it. “This means that restaurants can seamlessly raise or lower their prices based on the cost of supplies or even the prices of local competition,” Pittacora says.

New AI programs can help independent pizzerias with complex operational tasks such as order-taking, staffing and tracking inventory. Getty Images/Don Wu
A Little More Conversation

The rapid advance of AI can feel bewildering at times, but most of us are already familiar—and semi-comfortable—with conversational AI, which can field phone orders and book reservations. “We have all argued with a supposedly intelligent robot, finding that it only works if you know how to talk to it,” says John Scully, owner of PizzaCloud, a Granville, Ohio-based phone solutions provider that uses AI. “Conversational AI is different. You speak to the system exactly as you would a real person. You don’t have to say, ‘I want a LARGE…PEPPERONI…PIZZA.’ You can say, ‘Let me have a large pepperoni on thin crust…umm…add some black olives to that. No, make it an extra-large,’ and the AI gets it.”

But does it feel like a real conversation? “Well, the other day we had one caller apologize: ‘Hey, I’m sorry, I thought I was talking to a robot. But you’re a real person, right?’” Scully recalls. “The answer was, ‘I like to think of myself as an artificial person, but thank you.’”

Alex Sambvani, co-founder and CEO of Slang.ai in Brooklyn, New York, says booking a restaurant reservation with his company’s AI system “feels remarkably similar to speaking with a real person. [It’s] trained to understand and respond to customer inquiries in a natural and human-like manner.”

Every task that AI can perform is one less task a pizzeria owner has to pay an employee to do. “AI has tremendous potential to improve operational efficiency for smaller independent restaurateurs,” Sambvani says. “By automating routine tasks such as reservation management, order processing and customer service inquiries, AI frees up valuable time for restaurant staff to focus on delivering exceptional dining experiences.”

With Hot Rocks’ Volcano oven, computer vision can scan a pizza while it’s baking to ensure it meets your specifications every time.
Korok Group
Making Better Pizza

AI can also be used to make better pizza—which is something even tech-averse pizzaioli can appreciate. In 2019, Domino’s started testing an AI-powered system, called the DOM Pizza Checker, in Australia and New Zealand to make sure every pie meets the chain’s standards for quality and consistency. Now, thanks to Hot Rocks, a manufacturer of pizza ovens in Drummondville, Quebec, Canada, that kind of technology isn’t just for Domino’s franchisees Down Under.

Hot Rocks has equipped its Volcano oven with an AI camera from Kwali, a company that specializes in computer vision for foodservice and quick-service restaurants. Are you concerned that your pies are getting sauced or topped in a hurry or coming out burnt or underbaked? “It’s impossible to be in your stores checking every pizza before it goes into a box and out the door or to a table,” says Hot Rocks president Jason Bireta. “Often the owner or manager doesn’t know that low-quality product is being served until they see the negative review on social media. Subpar product will drive even loyal customers to your competition, either temporarily or forever.” 

With the Volcano oven’s AI camera and analytic capability, Bireta says, the operator receives a “snapshot report” on their pies. Does it have the right amount of toppings? Too much cheese? Not enough? “The system can tell that it’s actually a pizza coming out of the oven, what type of pizza it is—maybe a deluxe or pepperoni—and then grades it for you,” Bireta says. “You also receive a random sampling of product that comes out of the oven throughout the day or week.”

“We have all argued with a supposedly intelligent robot, finding that it only works if you know how to talk to it. Conversational AI is different.”

John Scully, PizzaCloud

The AI tech is a “sales and quality tool” that’s especially valuable for multiunit operators. “The report ranks how well a store or shift is performing, both over time and versus each other,” Bireta says. “So you can see the quality of pizza from store to store and from shift to shift. If you address the issues properly, your product quality will increase, ultimately increasing sales.”

With its ability to streamline operations, reduce overhead costs and boost productivity, AI can also make it easier to open your first pizza shop and turn a profit faster, Sambvani points out. “For example, AI-powered inventory management systems can optimize ingredient ordering, minimizing waste and controlling expenses. Additionally, AI-driven marketing solutions can help attract and retain customers, boosting revenue and profitability.”

For an AI beginner, the best first step is to become acquainted with a generative AI tool like ChatGPT, Pittacora says. “While it can’t help to automate tasks or tell you anything specific about your operations or customers, it can be a great way to spur ideas and create copy for signage, menu descriptions or more.” Try using it to create a list of specials, promotions or events to build traffic, she recommends, and go from there.

Sambvani predicts AI “will continue to play a significant role in transforming the restaurant sector by enhancing operational efficiency, improving guest experiences and driving revenue growth. We’ll see increased adoption of AI-powered solutions for tasks such as customer service, order management and personalized marketing. In the long term, AI has the potential to revolutionize every aspect of the restaurant industry, from kitchen operations to supply chain management, ushering in a new era of innovation and efficiency.”

Just remember: Aside from brainstorming, AI can’t do much for you without data. Get your data together first, and then get ready for one epic odyssey into the future.   

Rick Hynum is PMQ’s editor in chief.

Conversation Starters

For independent pizzeria operators, phone solutions offer an easy and useful way to start tapping into AI’s potential. John Scully, owner of PizzaCloud, recommends asking any vendor the following questions:

  • Can your solution be directly integrated with my point-of-sales system? (“If not,” Scully says, “you will likely experience, let’s say, high frustration.”)
  • Can the vendor properly handle delivery-address validation via your POS system? (“It’s critical to avoid AI taking delivery orders that you cannot deliver,” Scully notes.)
  • Can the vendor handle real-time pricing and availability updates, or will you have to call them whenever something changes?
  • Does the vendor have existing pizzeria clients that use your POS system?
  • Will the vendor give you a live demo? Recorded calls don’t count, Scully says. “Many vendors will happily send you what they claim are random recorded calls from real people. But if you ask for a number that you can call to play with the system, the response is, ‘Umm…let me get back to you.’ If you cannot test it, does it really work?”
Featured, Technology