As labor costs rise and customer expectations evolve, QSRs are increasingly adopting voice AI technology to streamline operations and improve the experience for both guests and employees.
The technology automates the ordering process, freeing up staff for other tasks while offering operators a tool to tackle labor challenges. Beyond that, tech vendors often highlight AI’s potential to boost sales through automated upselling—an area where human employees, especially after long shifts, may be inconsistent.
Quick-service customers are increasingly encountering voice AI at the drive-thru, but it’s also making a splash over the phone, offering a much-needed improvement over the frustrating automated systems of the past.
MORE: The 2025 Crystal Ball for Restaurant Tech: It Doesn’t End with AI
“Today, AI is at the point where we’re able to create a very clean, frictionless, uninterrupted experience for the guest,” says Vinay Shukla, cofounder and CEO of ConverseNow, a voice AI technology platform for restaurants. “It engages with the customer and naturally navigates with them throughout the conversation—discussing the menu, answering questions, and validating the delivery address. It’s able to take complex orders and it’s able to get more upsell.”
When ConverseNow first launched its voice AI in 2019, both operators and customers were hesitant about speaking to a machine instead of a live person, he adds. But times have changed. The rise of ChatGPT and other AI tools has boosted awareness and made consumers more comfortable with the technology. As people become more familiar with AI, their willingness to interact with it has grown significantly.
And it’s not just customer attitudes that have evolved—the technology itself is now far more advanced. Early voice AI relied on rigid, rules-based systems that mimicked a flowchart, handling conversations with fixed logic. Today’s large language models (LLMs) are a game changer. Instead of following strict rules, LLMs are trained on massive datasets, allowing them to understand and generate complex language. This flexibility means they can respond more naturally, making interactions smoother and more personalized.
Thanks to this sophistication, voice AI can now manage more complicated interactions. Shukla points to phone orders as an example. Customers often don’t have a full menu in front of them and ask lots of questions about meal sizes, discounts, specific ingredients, and the delivery or pickup process. With AI, these inquiries are handled effortlessly, guiding the customer through the process in a way that once required human intervention.
For restaurants considering voice AI, Shukla has a few tips. First, make sure your POS system can handle AI integration—many older systems don’t have the necessary APIs to work smoothly with newer technology. Second, don’t just cut staff after adopting AI. Instead, reassign them to roles that enhance the guest experience. Finally, remember that some customers will still prefer talking to a human, so don’t expect to immediately rely solely on AI.
Shukla also emphasizes the importance of educating staff on what to expect from AI and allowing customers to adjust to the technology at their own pace, rather than forcing it on them. Allen Beck, director of off-premises and catering at Costa Vida, notes that patience is key during the implementation phase.
“When you start, you have to understand that there’s a learning curve, just like there is when a human comes into the call center,” he says. “Just because it’s this awesome artificial intelligence with a complex algorithm doesn’t mean that it knows your business. But the more calls that come in, the more it’s learning, and the more questions it can answer.”
Costa Vida rolled out voice AI in its call center earlier this year. Staffing shortages drove the decision—hiring and training call center agents in Utah, where the company is based, has been increasingly difficult due to competition from other call centers. Voice AI presented a solution to that problem that could also help reduce labor costs.
Roughly 75 percent of the calls coming into Costa Vida’s call center were generic inquiries about things like store hours, drive-thru availability, patio seating, and food allergens. Beck recognized these as prime opportunities for AI to step in. By offloading the frequently asked questions, Costa Vida was able to reduce the number of human agents required.
Over time, the system has become more adept at managing complex interactions.
“Before, it was the basics,” Beck says. “Now, it’s adapted and it can do more. If someone calls in and says they’d like to apply to work at this location, the AI confirms the phone number and shoots them a text message with the application. That wasn’t possible six months ago.”
Looking ahead, Costa Vida plans to deepen its investment in voice AI. The plan is to expand its use in the call center by fully automating the ordering process. Beck sees this as a way to improve accuracy while further cutting down on labor costs, but he still advises operators to approach AI adoption thoughtfully.
“AI is such a hot buzzword right now,” he says. “So many people want to get into it, but they don’t know what they want to get into it for. Just like with anything else, you want to make sure you know what you want to do before you go out and start looking for solutions.”
Austin Titus, president of Cannoli Kitchen Pizza, echoes this cautious approach when evaluating voice AI for restaurants. He advises restaurants to carefully weigh their options before diving in. Titus hasn’t implemented AI for phone orders yet, but he’s done his homework—researching the technology extensively and speaking with vendors at industry events.
“Busy pizzerias that do a lot of delivery have to have one or sometimes even two full-time staff on a Friday and Saturday night just to answer the phones,” Titus explains. “It’s a lot of man hours that have to be put in for something that’s a very simple task.”
At some Cannoli Kitchen Pizza locations, anywhere from 40 to 70 percent of orders come in by phone, which makes the potential for AI significant. But Titus remains cautious, knowing that one bad experience can lead a customer to hang up and call a competitor.
“I always say to walk and not run to these things,” he says. “They could blow up your business in a great way, but they could also blow up your business in a bad way if you’re not careful.”
