Restaurant AI: Why operators need answers, not more dashboards
Restaurant operators have never had more data. Sales dashboards, labour reports, inventory systems, forecasting tools. And yet, ask most operators whether their software actually makes their decisions easier, and the answer is a polite no.
That gap is where the next wave of restaurant AI is being built. Nory’s founder and CEO, Conor Sheridan sat down with Shawn Walchef on the Digital Hospitality podcast recently to walk through exactly what that shift looks like in practice.
The dashboard problem
When Conor left a career in quantitative finance to open Mad Egg (his fried-chicken concept in Dublin), he was coming from an industry where technology made decisions at breakneck speed. Trading systems didn't just show you the market. They told you what to do with it.
Restaurant software didn't work that way. Point of sale, scheduling, inventory, business intelligence, procurement, Conor tested nearly every category while scaling Mad Egg. All of them helped digitise workflows. None of them actually did the work.
"They gave you visibility," Conor told Shawn. "They didn't do the work on behalf of the teams."
The result was familiar to anyone who has scaled a multi-site operation. Managers were juggling labour, purchasing, prep, forecasting, and inventory while trying to deliver a great guest experience. Even with the same menu, similar traffic patterns, and standardised operating procedures, sites performed differently. The variability came from the operational decisions being made unevenly across the estate. Not from the brand, not from the concept, not from the customer.
What Agentic AI actually looks like in restaurants
The word "AI" has been getting applied to restaurant technology fast enough to trigger fatigue in most operators. Conor's frame on Digital Hospitality is more specific: agentic AI is defined by autonomy, not by conversation.
The distinction matters. A chatbot that can answer questions about your P&L is useful. A system that spots a shift running behind forecast, recalculates the purchase order, adjusts the prep list, updates the staffing plan, and surfaces the whole thing for the manager to approve is a different category of product.
That's the direction Nory has been building toward. Not a smarter dashboard. Software that takes the decisions the manager was going to have to make anyway, and hands them the plan.
Conor's shorthand for it: "People don't want insights. They want answers."
Why this matters now
The economics of the restaurant industry have compressed on both sides of the Atlantic over the last five years. Labour costs, food costs, aggregator commissions, business rates. The margin cushion that used to absorb operational variability isn't there any more.
For an operator scaling from one site to ten, or ten to a hundred, small operational mistakes get expensive fast. Missed forecasts, over-prep, understaffed shifts, purchasing errors, each one individually manageable, collectively a slow leak on the P&L.
Software that reports on those problems after they have happened isn't the answer. Software that takes action while there is still time to fix them is.
Listen to the full conversation
The full episode covers agentic AI in restaurants, the US market entry, the Toast integration story, and the origin of the name Nory (which is not what most people assume).
Or explore more operator conversations on What's Cooking, Nory's podcast.