What AI Can Actually Do for Your Restaurant — And Why the Conversation Starts with Curiosity
A few years ago, I started noticing something in my conversations with restaurant operators.
The ones getting the most out of new tools weren’t the most tech-savvy. They weren’t the youngest, or the ones with the biggest budgets. They were the ones who stayed curious — who asked the next question, pushed on the answer, and kept going when the first result wasn’t quite right.
That observation is what eventually led to Urban Food AI. And it’s the right place to start this blog.
What This Blog Is For
This is a space for restaurant operators who want to think more clearly about their businesses. Not a place for hype, and not a pitch for any particular tool or service. Every post here will share something I’ve learned — from thirty years working with operators in the NY Metro area, from culinary fieldwork across 15 countries, or from the AI work I’ve been doing since completing the Generative AI for Business Transformation program at Cornell Tech.
Some posts will be practical. Some will be philosophical. All of them will be honest.
Rethinking What AI Is
The most useful shift I’ve seen operators make is a simple one: stop thinking of AI as a tool that produces answers and start thinking of it as a partner that helps you ask better questions.
When I work through a business review with an operator using AI-assisted analysis, we’re not just pulling numbers. We’re building a story — where the business has been, what the data is actually saying, and where the clearest opportunities are. That process has consistently cut preparation and analysis time by around 70% compared to doing it manually. More importantly, it changes the quality of the conversation. Operators come out of it seeing their business differently.
That shift — from AI as a shortcut to AI as a co-intelligent partner — is the one worth making.
The Skill That Makes the Difference
If there’s one thing I’d encourage every operator to develop right now, it’s the ability to ask good questions of their data and their tools.
The technical term for this in AI is prompt craft — how you frame a question shapes what you get back. But it’s really just curiosity applied deliberately. The operators I’ve seen get the most from AI are the ones who don’t accept the first answer, who try a different angle, who bring their own knowledge into the conversation and push back when something doesn’t feel right.
This is learnable. It doesn’t require a technical background. It just requires a willingness to stay in the conversation.
The Framework Behind the Work
One idea runs through everything I do with operators, and it’s worth naming here because it will come up again in future posts.
Start from where you want to be. Then work backwards to what that requires today.
Before signing a lease, describe the business you want in five years — then ask whether this location gets you there. Before adding a menu item, define what success looks like on the P&L — then figure out what the recipe needs to be to hit that number. Before opening a second location, decide what it needs to prove about the long-term vision.
This is how the most effective operators I’ve worked with naturally think. It turns out it’s also how well-designed AI systems reason — they’re built to work backwards from a desired outcome. The two approaches fit together naturally, which is part of why AI can be genuinely useful for operators who already think this way, and why learning to think this way makes AI more useful.
What’s Coming on This Blog
In the posts ahead, I’ll dig into the things that actually move the needle for operators:
What your online reviews are telling you that you’re probably not hearing
How to read a menu as a P&L — and what that reveals about where your margin is going
What I’ve learned from culinary fieldwork across Italy, Morocco, the Republic of Georgia, Mexico, Colombia, and more — and what it means for NY Metro menus
Where AI genuinely helps in restaurant operations, and where the claims outpace the reality
Site selection, lease strategy, delivery channel optimization, and the invisible problems that tend to show up only after they’ve already cost something
The goal is always the same: to share something useful. Something you can think about on the subway, bring into a conversation with your team, or apply directly to a decision you’re facing.
A Note on How I Work
Urban Food AI grew out of thirty years of working alongside independent operators in the NY Metro area — helping them think through expansions, engineer menus, read their markets, and navigate the decisions that don’t come with obvious answers.
The AI tools came later, but they fit naturally into that same kind of work. They’re most useful when they’re combined with real operator knowledge — someone who understands what the data means when it comes back, and can help translate it into something actionable.
That’s what this blog is an extension of. Not a platform, not a product. Just a practice of paying attention and sharing what’s worth knowing.
If something here is useful to you, I’m glad. And if you want to talk through something specific about your restaurant, the door is always open.
Like this? Get new posts by email — no hype, no pitch. Scroll up to the Blog page and add your address under “Get New Posts by Email.”
