Applebee’s and IHOP’s CIO places AI bets that can boost traffic and improve the dining experience

This AI-powered “recommendation engine,” which utilizes Google Cloud’s recommendations AI technology, is already in production for online orders at the restaurant operator’s IHOP and Applebee’s chains and will also roll out soon at Dine’s Fuzzy’s Taco Shop. It blends together a guest’s past order history with the buying patterns from a broader pool of similar diners to generate more personalized menu pairings.

But there could be even greater opportunities for targeted, in-person upselling when the technology could be deployed across the chain’s 3,5000 brick-and-mortar restaurants.

“There’s a lot of new technologies we’re looking at that we want to bring to the market,” says Skelton, who has served as CIO at Dine since 2019, after prior technology leadership roles at CVS Health, Bank of America, and Allstate.

Skelton says that all of his generative AI bets are intended to support three outcomes: boost revenue and traffic, improve the guest experience, and bolster worker productivity. Because he runs a small IT team, Skelton says he has to be very careful about his technology bets. “We’re not prepared to go and invest in and hire a lot of people,” says Skelton.

With that in mind, Dine tends to partner and “buy” more solutions from vendors rather than build. Skelton relies on large AI companies like Google and Amazon and non-hyperscalers like IT consulting firm Cognizant. 

The latter helped Dine create and launch an AI Innovation Foundry last fall, a virtual lab that is led by Skelton’s IT team and works with different stakeholder groups including legal and finance to test emerging AI technologies. Initially only available for corporate employees, the Foundry became so appetizing that franchise partners have also been encouraged to submit use cases for Dine’s team to vet.

“There’s a recognition that not all ideas are going to make it and see the light of day,” says Skelton. “It’s really sort of a way we can incubate.” 

With budget planning for 2026 already under way, Skelton and his team will mull the proposed use cases and a steering committee will determine the few select projects that will get capital allocation.

Some newer concepts that Dine is considering include AI-enabled tools that can make it easier for the general manager to understand their staffing levels, inventory, and other data needed to make a restaurant run smoothly and utilize that real-time data to make better decisions.

Another is an evolution of Dine’s generative AI assistant tool that guides the company’s franchise technology services team, a group of technicians that resolves issues like when a printer breaks or a point of sale system is down at a restaurant. In the future, Dine hopes to make this a self-service tool, allowing franchisees to access Dine’s technician data to solve some of their own problems.

Investments in AI to improve back office functions, demand forecasting, and predicting the appropriate staffing level for any given day are all efforts that can bolster the bottom line. That’s especially critical in the restaurant industry, which operates on notoriously thin margins.

For customer-facing applications of AI, many restaurant chains have prioritized AI-enabled recommendation tools that are flourishing online. Chains like Taco Bell, McDonald’s, and Wendy’s have all deployed or tested AI-enabled drive-thru technologies to varying degrees of success.

There is also a fine line that casual-dining chains like Dine must balance when adding technology—a playbook that differs from fast-food operators like McDonald’s and Burger King. At those chains, guests often prioritize the speed of service, which has fueled the popularity of ordering kiosks in recent years. But at Applebee’s or IHOP, diners crave a more personal connection.

“There’s still that human touch and that human interaction that cannot be replaced with technology,” says Skelton. “But it can be enhanced, without a question, and we’ve got to figure out how to get that right.”

John Kell

Send thoughts or suggestions to CIO Intelligence here.

NEWS PACKETS

The AI talent war continues to intensify. A lot of new media reports have homed in on the aggressively expensive AI talent war, including a piece from The Wall Street Journal that cites research showing that Anthropic’s retention rate is higher than rivals OpenAI, Meta, and Google. The Financial Times reports that Apple has been losing AI staff to rivals throughout the year while Business Insider says that Elon Musk’s xAI has hired 14 AI researchers and engineers from Meta since January. OpenAI CEO Sam Altman, commenting on the marketplace in a recent interview with CNBC, says this is “the most intense talent market I have seen in my career.” He adds that the compensation packages will be especially intense for “a very small number of people right now.” The New York Times reports that pay packages for the most elite talent is rivaling that of NBA stars like LeBron James and Steph Curry.

GPT-5's bumpy debut. Last week, OpenAI launched a new version of its flagship product, GPT-5, which the company says delivers “more accurate answers than any previous reasoning model,” and is less prone to hallucinations than previous models, has better coding functionality, scores higher to offer medical health advice, and integrates with Gmail and Google Calendar for agentic capabilities. But the product's "real-time model router"—which automatically decides which one of GPT-5’s several variants is best suited for each particular job—sparked a user backlash. While the router was designed to save users from the hassle of manually pickings models, many users bristled at the lack of control. Still, as Fortune's Sharon Goldman reports, these model routers may be the future of AI. 

Nvidia, AMD strike deal for export licenses in China. American chip makers Nvidia and Advanced Micro Devices (AMD is the sponsor of this newsletter) have agreed to pay 15% of their revenue from Chinese AI chip sales to the U.S. government, an unusual arrangement that Bloomberg reports may set a new precedent for U.S. companies that seek to do business in China. Nvidia announced its plans to share 15% of the revenue of the company’s H20 AI accelerator in China, according to President Donald Trump in a briefing with reporters on Monday, while AMD’s deal was reported by Bloomberg, citing a person familiar with the matter. Separately on the chips front, Trump last week demanded the resignation of Intel CEO Lip-Bu Tan, following a request from Republican Sen. Tom Cotton for more information about Tan’s ties to Chinese companies. After Tan met with Trump on Monday though, the POTUS said he would be spending more time working things out with Tan, whose career he pronounced an "amazing story."

ADOPTION CURVE

Nearly half of developers don’t trust the accuracy of AI tools. An international survey of over 49,000 developers found that a lack of trust is growing in AI tools even as the use of these tools proliferates in the development process. 46% of developers say they don’t trust the accuracy of the output of AI tools, a big increase from 31% in last year’s survey, which was conducted by developer question-and-answer website Stack Overflow.

Other notable nuggets from the study include that a majority of developers still aren’t using AI agents (only 31% are doing so). The large language models that are used by most developers are OpenAI’s GPT models (81%), Anthropic’s Claude Sonnet models (43%), and Google’s Gemini Flash models (35%).

When it comes to technologists and their evolving approach to work, 66% of developers say they spent time in the last year learning new coding techniques or a new programming language and 44% have turned to AI tools to learn how to code, up from 37% last year. But “vibe coding,” a buzzy new practice of writing code with AI using natural language prompts, still isn’t embraced by most. Nearly 77% say they aren’t vibe coding for their professional work.

Courtesy of Stack Overflow

JOBS RADAR

Hiring:

- Datadog is seeking a field CTO, with a focus on the financial services industry, based in New York City. Posted salary range: $280K-$380K/year.

- Citi is seeking a managing director, head of shared services technology, based in Jersey City, New Jersey. Posted salary range: $250K-$500K/year.

- S&P Global is seeking a head of technology, market analytics, based in New York City. Posted salary range: $185K-$300K/year.

- American Express is seeking a director of digital program management, based in New York City. Posted salary range: $150K-$225K/year.

Hired:

- SpartanNash (No. 428 on the Fortune 500) announced two new IT leaders to join the food distributor and grocery store retailer, naming Ed Rybicki as SVP and CIO and Brett Hoffman as VP and chief information security officer. Rybicki most recently served as chief information and technology officer at agricultural producer and supplier Mastronardi Produce, while Hoffman joined SpartanNash from security and risk management consultancy Inspire Security Solutions, where he served as CISO.

Every Friday morning, the weekly Fortune 500 Power Moves column tracks Fortune 500 companies C-suite shifts—see the most recent edition.

- H&R Block announced the appointment of Jason Lenhart as SVP and CTO, joining the tax-preparation company after most recently serving as VP of technology engineering for airline JetBlue, where he led software development. He also held technology leadership roles at HBO Max, Barclays, and Comcast.

- Upwork has expanded the role of Andrew Rabinovich, who has taken on the CTO responsibilities on top of his work leading AI and machine learning. Rabinovich will now lead engineering and IT and since joining the freelancer marketplace in late 2023 as part of its acquisition of Headroom, has overseen the development and evolution of the AI agent Uma. Rabinovich previously served as head of AI at augmented reality company Magic Leap.

- Ziosk named Jacob Feinstein as CTO, effective immediately, to oversee the restaurant technology provider’s engineering organization. Previously, Feinstein spent nearly three decades at telecommunications giant AT&T, where he most recently served as vice president of digital experience and product realization.

- Pathward Financial has promoted Charles Ingram to serve as EVP and chief information and operations officer of the financial services company, which he initially joined as CIO in 2020. He also served as EVP and chief technology and product officer for four years. Prior to his time at Pathward, Ingram held technology leadership roles at retailer PetSmart and restaurant operator Yum Brands.

- Greenhouse announced the appointment of Sagar Patel as CTO, joining the hiring software company after most recently serving as head of engineering for revolving credit products at financial technology company PayPal. He also previously served as CTO of financial technology provider Ampla and was a VP of engineering at asset manager BlackRock.

- Terrepower appointed Richard Longsdorf as EVP and CIO, following the planned retirement of Tom Sheppard, who will transition to an internal advisory role until the end of the year to ensure a smooth handover. Longsdorf joins the automotive and industrial products manufacturer from supply chain management company Fidelitone, where he served as CIO.

- AWP Safety named Craig Young as CIO, joining the provider of outsourced traffic control workers to oversee digital integration, automation, and data transparency as the industry adapts to labor shortages. Young previously served as CTO at fiber-optic telecommunications company Tillman FiberCo and as global CIO at financial messaging provider Swift.

- Weave promoted Abhi Sharma to the role of CTO, after initially joining the payments software provider in March as SVP and head of engineering. Prior to joining Weave, Sharma served as SVP at software company Twilio and as a VP at cloud software provider Salesforce. Earlier in his career, Sharma held technology leadership roles at Oracle, Amazon, and Microsoft.