By Lea Mira and Orit Naomi, RTN staff writers - 8.1.2026
For all the restaurant industry’s excitement about artificial intelligence, one of the most consequential technology decisions at Chili’s was replacing its Wi-Fi access points. That work began after Chris Caldwell joined parent company Brinker International as chief information officer in February 2024. Caldwell found restaurant managers and servers struggling with unstable connectivity, aging devices and applications that made routine tasks unnecessarily difficult.
Recent articles in The Wall Street Journal and Fortune detailed the resulting overhaul. Instead of making generative AI the centerpiece of Brinker’s technology program, Caldwell first concentrated on the systems employees use during every shift. Caldwell brought considerable restaurant technology experience to the assignment. Before joining Brinker, he spent 27 years at Yum! Brands, most recently leading technology for KFC in the United States. Brinker CEO Kevin Hochman, another former Yum executive, had already concluded that technology problems were making restaurant work harder than necessary. Fortune reported that as much as 70% of the employee feedback Hochman reviewed after joining Brinker involved technology-related frustrations.
The first major project was a network rebuild covering approximately 1,200 Chili’s and Maggiano’s Little Italy restaurants. Brinker retained Comcast Business, installed new networking equipment, replaced the wireless access points, added cellular backup and brought fiber connections to some restaurants where existing service was inadequate. The phased project took about two years and was completed earlier this year.
Reliable connectivity opened the door to a wider device refresh. Brinker supplied approximately 1,200 laptops so restaurant managers could handle inventory, email and administrative work without relying entirely on overloaded back-office computers. The company also deployed about 23,000 new Apple iPads to replace server tablets whose batteries and performance had deteriorated.
Another 9,000 touchscreens were installed or reconfigured in restaurant kitchens. Instead of showing employees a crowded list of every open order, the screens present the work relevant to each kitchen station, making it easier for cooks to see what they need to prepare next. Brinker also began redesigning the handheld ordering application used by servers. The new interface was being tested in 12 restaurants when Fortune reported on the project in June. Brinker’s goal was to cut the number of screen taps required to enter an order by half and make the application intuitive enough for a new employee to learn with limited training.
The company also returned to Ziosk as Chili’s tabletop payment provider in 2024, replacing the Presto platform it had used since 2020. The rollout brought Ziosk devices to more than 1,100 company-owned restaurants, allowing guests to pay without waiting for a server while also supporting My Chili’s Rewards enrollment, entertainment and post-visit feedback. The move fits Caldwell’s emphasis on practical technology that removes friction from the dining experience rather than introducing automation for its own sake.
The kitchen work builds on Brinker’s long relationship with QSR Automations by Crunchtime. As Restaurant Technology News reported in 2023, Brinker renewed that partnership as Chili’s prepared to upgrade its kitchen display system to ConnectSmart Kitchen. The relationship dates to 2001, when Chili’s became an early casual-dining adopter of the vendor’s kitchen technology.
None of these projects carries the novelty of a restaurant robot or an AI ordering agent. Their value is easier to see from the restaurant floor, where a dead tablet battery, a dropped network connection or a confusing order screen can slow service across several tables at once. A few seconds lost to an unresponsive device may appear insignificant when measured as an isolated IT incident. Repeated across thousands of employees and millions of annual transactions, those delays affect order accuracy, table turns, employee frustration and the amount of attention servers can give guests.
Brinker’s investment has coincided with a striking recovery at Chili’s, although technology is only one part of the story. The company has also simplified its menu, strengthened its value proposition, improved food and service, increased marketing and benefited from renewed interest in products such as the Triple Dipper. In the third quarter of fiscal 2026, Chili’s reported its 20th consecutive quarter of comparable restaurant sales growth. Comparable sales increased 4%, despite comparison with a 31% gain in the prior-year quarter. Sales in February and March each rose 5.9% with positive traffic, according to Brinker.
Brinker attributed the performance to improvements in food, service and atmosphere, supported by menu innovation, everyday value and advertising. The technology program has supported that strategy by removing obstacles that previously made it harder for restaurant teams to execute it consistently. The decision to concentrate on basic systems also reflects lessons from Chili’s earlier experiments with more visible automation. In 2022, the chain paused the expansion of Rita, a service robot from Bear Robotics that had been deployed in 61 restaurants.
Rita could carry food, bus dishes, guide guests to tables and perform novelty functions such as singing birthday songs. Brinker said the test produced encouraging results, but management did not see a sufficiently clear path to a financial return and decided to direct resources toward projects with a more immediate effect on restaurant margins and operations.
The company’s current position is not that robots or AI have no place at Chili’s. Caldwell’s standard is whether the technology improves the experience for guests and restaurant employees. A system that creates congestion, adds work or distracts from hospitality is unlikely to survive simply because it uses a fashionable technology.
Brinker is applying the same filter to generative AI. The leadership team recently considered several dozen possible applications, but Caldwell narrowed the list to six or seven ideas worth further investigation. A governance group now reviews proposed AI projects monthly and reconsiders previously rejected uses as the technology develops. Inventory forecasting and replenishment is one of the applications moving forward. Better forecasting could help managers decide how much product to order, reduce administrative work and eventually automate parts of the replenishment process.
Brinker has been less enthusiastic about using AI to answer telephone orders. Caldwell concluded that the available systems were more likely to frustrate guests than improve their experience, according to the Journal. The company may revisit the idea as the technology improves, but it is not deploying the application simply to demonstrate that Chili’s has an AI program.
That measured approach differs from the more expansive AI strategies being pursued by some of the industry’s largest restaurant companies. Yum! Brands has consolidated many of its restaurant systems into Byte by Yum, a proprietary platform covering online ordering, point of sale, kitchen operations, delivery optimization, menu management, inventory, labor and employee tools. Yum reported in its 2025 annual review that more than 38,000 restaurants were using at least one Byte product.
Yum’s approach gives it a common technology layer across KFC, Taco Bell, Pizza Hut and Habit Burger & Grill. That scale allows the company to introduce AI for forecasting, restaurant coaching and other functions using data generated by systems it already owns or controls. Chili’s is pursuing some of the same operational goals without trying to build an equivalent global platform. Brinker has first concentrated on making its networks and restaurant devices dependable, then evaluating AI where it can improve a specific task.
McDonald’s is also treating infrastructure as a prerequisite for restaurant AI. Its Restaurant Platform Edge, developed with Google Cloud, brings computing capacity into individual restaurants and is operating in hundreds of U.S. locations. McDonald’s describes the platform as a foundation for AI and connected-equipment applications intended to improve uptime, food quality and restaurant reliability. McDonald’s has paired that foundation with narrower AI applications. Its Accuracy Scales compare the expected weight of an outgoing order with the actual weight and alert employees when an item may be missing. The company said in 2025 that the scales had been installed in thousands of restaurants across a dozen markets.
Chipotle Mexican Grill has concentrated much of its automation work on repetitive food-preparation and assembly tasks. Its Autocado prototype, developed with Vebu, cuts, cores and peels avocados before employees mash them by hand. An augmented makeline created with Hyphen automatically assembles bowls and salads while employees use the upper portion of the line to prepare burritos, tacos and other items. Those projects reflect a shared pattern across the more credible restaurant automation programs. The technology is assigned to a defined job with a result that operators can measure, such as fewer missing items, less preparation work, faster order entry or more accurate inventory forecasts.
Chili’s faces a somewhat different operating challenge because casual dining depends heavily on the interaction between servers and guests. The objective is rarely to remove the employee from the dining experience entirely. Technology earns its place when it gives servers more time to manage tables, answer questions and provide hospitality.
A reliable handheld can support that goal by allowing an order to reach the kitchen immediately. A well-designed kitchen screen helps cooks coordinate the meal, while a functioning tabletop payment device prevents the guest from waiting after the check arrives. Those improvements may sound modest beside a generative AI assistant, but they reach nearly every customer and employee. They also create cleaner operational data and more dependable connections for future applications.
Technical debt remains a widespread problem in restaurants. Companies frequently add ordering channels, payment devices, loyalty systems and employee applications without retiring older technology. The result can be a patchwork of software, hardware and credentials that asks restaurant employees to compensate for gaps between systems. Adding AI to that environment can make the stack more complicated without fixing the original problem. A forecasting model cannot produce reliable recommendations when the underlying sales, inventory or menu data is inconsistent. A digital employee assistant will not improve service when the restaurant network cannot keep the device connected.
Brinker’s sequencing reduces that risk. The company rebuilt the network, refreshed the devices and began simplifying the applications before expanding its AI program. The approach does not require Chili’s to wait until every legacy system has been replaced. It does require each new project to solve an identifiable restaurant problem and work within the operating environment employees face during a busy shift.
Caldwell’s governance group should help maintain that discipline. Restaurant companies are being approached by a growing number of vendors promising labor savings, faster service and better decisions through AI. A recurring review gives Brinker a way to compare those claims with actual operating priorities and stop projects that fail to produce useful results. The rebuilt technology foundation should also give Chili’s more options. Inventory forecasting is the first publicly identified AI application, but the same infrastructure could eventually support carefully selected work in maintenance, preparation timing, workforce planning and guest service. Those opportunities will depend on the quality of Brinker’s data and whether each application produces results that restaurant teams can use.
Chili’s is not withdrawing from restaurant innovation. It is insisting that innovation perform useful work. That position may become more common as the industry moves beyond the first wave of generative AI announcements. Restaurant operators will increasingly judge technology by order accuracy, throughput, uptime, employee adoption and guest satisfaction rather than the sophistication of the demonstration. For Chili’s, the technology turnaround began with the tools employees already use. The company now has a stronger foundation for AI, but only where AI can make the restaurant work better.
