Several Houston restaurants have cracked the code on efficiency, using artificial intelligence to streamline small tasks and free up staff to focus on hospitality.
ChatGPT and other AI tools have surged in popularity across industries, and the restaurant world is no exception. From fast-casual spots to fine dining, local operators are finding ways to integrate the technology. Much of it happens behind the scenes, invisible to customers, though some restaurants are beginning to explore guest-facing uses, as well.
“There is constant stress on the restaurant operation business model. So if you’re able to implement the right artificial intelligence, either to enhance your guest experience or to help automate some of those roles that are not hospitality-based, then for the future of the industry, that is something that you have to embrace,” said Emily Knight, president of the Texas Restaurant Association.
[…]
Other AI companies are rolling out tech specifically to help restaurants forecast their prep portions and costs. Mac Haik Restaurant Group — which owns First Watch, Slapfish, the Original Chopshop and Due Cucina — is using Prep Assistant from Precitaste, which predicts how much food a restaurant will need each day and creates a prep list that reduces waste. It even factors in weather and other events that might affect customer traffic.
Training that once took managers six months on the old inventory system now takes about a week with Prep Assistant. Prep work is finished faster, allowing staff to spend more time on hospitality. The shift has boosted sales and, in turn, increased employee earnings through tips, said Chief Operating Officer Dan Anfinson.
The company has seen at least a 50% reduction in food waste for two consecutive fiscal periods. “The system has been more accurate for what I need, what we need to prep, and the shelf life of food,” said First Watch manager Araceli Salgado. “We don’t have to throw out batches of items because we over-prepped. It’s helping with training, knowledge, and the quality of both the food and production.”
Prep Assistant runs on a subscription model, but Anfinson said it has already paid for itself in the month since its roll-out across the group. He added that the company worked with PreciTaste for more than a year to tailor the system to their needs.
That’s one of three examples cited – the first is a mostly LLM example, and the third is an automated voice system that seems to build on existing automated voice tech systems. Neither it nor this one here would fall under what I think of as ChatGPT-type systems, but somewhat surprisingly to me they all sound like reasonable use cases. None seem like direct threats to jobs, either, though the first example uses a tool that would reduce the need for an accountant. The best uses of AI, or more generally automation that may or may not be AI-based, is to handle low-value repetitive tasks and allow the humans to do more high-value work. If that’s the way restaurants are deploying this now, then good for them.
(The embedded image is from the site AI Darwin Awards, which aims to honor the least best uses of AI from the past year. I very much look forward to seeing who the winners are.)