Can AI Design a Commercial Kitchen?

Technically, yes. But that’s not actually the question worth asking.

The question worth asking is whether it can design one well. And that’s where things get interesting… or, depending on where you sit in the industry, where they get heated.

This is the first in a short series on AI in commercial foodservice covering design, operations, and environmental impact.

There’s a particular conversation happening in foodservice at the moment. Mention AI and watch as the room sorts itself into two camps: cautious curiosity on one side, and on the other, an immediate resistance. Usually driven by the same concern: will it take our jobs?

Specifically in this case, will it be able to design a commercial kitchen? Which, on the surface, sounds plausible.

So let’s look at it properly.

Where it starts: understanding the client

Good kitchen design doesn’t begin at a drawing board. It begins with people. A series of revealing, sometimes awkward, conversations in which you build up a picture not just of what someone needs, but of what they actually want, and what they haven’t yet found the words for.

That’s not something you get from a questionnaire.

The useful information you gather from conversations often isn’t from the words themselves. It’s in the slight tension when a particular layout comes up or the raised eyebrow when you mention a supplier they’ve had trouble with before. It’s in the long pause before someone says, “I suppose that could work” (which, in practice, means “please don’t do that.”)

It’s about gathering data whilst reading people to actually understand what that data means to each person.

AI cannot do this. Not because it isn’t clever, but because the information it needs isn’t in any format it can access. And while the technology will undoubtedly continue improving, technical capability and professional judgement are not necessarily the same thing.

The input problem

Here’s the part that gets less airtime than it deserves: AI is only as good as what you put into it.

And in a commercial kitchen project, the information that actually matters often isn’t neatly documented. It’s the drainage run that doesn’t appear on the plans. The wall that was altered years ago but never updated on the drawings. The awkward corner behind the servery that only becomes a problem once the site is operational. The bottleneck that appears every lunchtime despite the layout looking perfect on paper.

You could argue that a very thorough client questionnaire might capture some of it. But would a client know to include it? Would they know which details matter? Would they even know what they didn’t know?

The answer to all three is: not reliably, no.

Which means the quality of any AI-generated output is, in practice, dependent on the expertise of whoever is feeding it information. And who has that expertise? Experienced consultants.

Not because they have spent years on one site, solving one type of problem, but because they have worked across multiple sites, multiple operational models, multiple teams, multiple failures, and multiple scenarios. They have seen what happens when a design works brilliantly in theory but collapses under pressure at lunchtime. They have seen how different people behave in spaces, how staffing changes affect flow, how culture affects operations, and how small design decisions ripple into service, waste, atmosphere, and profitability.

That breadth of experience is what gives context to the information being fed into AI. Without it, AI can produce something convincing. With it, it can become genuinely useful.

What happens when things go wrong

Projects can go wrong. Part of the value of having a consultant on a project isn’t just the design itself. It’s the problem-solving that happens as the project is going in, the ability to adapt when reality diverges from the drawing.

There’s also the matter of accountability. When something does go wrong – and in commercial catering, the stakes aren’t abstract; they’re food safety, operational efficiency, people’s livelihoods, someone has to be responsible.

A consultant carries that. AI does not, and cannot.

The risk has to sit somewhere. Worth considering where, before committing to who does the design.

What AI can actually do

With the right information, and the right expertise shaping that information, AI can produce workable layouts. It can iterate quickly. It can surface options that a human might not have considered, or might have taken longer to reach.

These are not trivial things.

But “workable” is not the same as “good”.

And in commercial kitchen design, where every site has its own constraints, its own customer profile, its own way of working – mediocrity has real consequences. It shows up in productivity, in throughput, and in the daily experience of the people who work in that space.

The consultants who exist are there for a reason. Each kitchen is different. Each space is different. Expertise is the variable that turns a layout into a kitchen that actually functions.

It’s also important not to collapse every AI conversation into the same category. AI-generated design is one subject. AI-enabled equipment, monitoring systems, operational analytics, predictive maintenance, energy management, and data-led kitchen operations are another entirely and arguably the area where the most immediate change is happening.

There’s already some interesting discussion within FCSI around the growing role of AI in professional kitchen design and operations, particularly around data, optimisation, and systems integration.

Consultants understanding how those technologies can improve operations, sustainability, staffing, maintenance, and decision-making is not the same discussion as whether AI can independently replace the process of designing a kitchen in the first place.

That is a separate conversation, and frankly, a much more interesting one. I’ll come back to it in the next article.

The three places people will land

Here’s where I think the industry is heading, and it’s probably less dramatic than the conversation suggests.

Those using AI without the expertise to underpin it will likely produce mediocre results. Fast, dangerously convincing, and potentially useful in certain situations, but rarely enough for the complexity of a real commercial kitchen project.

Those with deep operational and design expertise will remain entirely relevant, whether they choose to actively use AI tools or not. Hospitality has always relied on judgement, interpretation, and experience. In many ways, the growing popularity of open kitchens reflects that. People increasingly value visible craft, human decision-making, and the reassurance that there is real thought and skill behind what they’re experiencing.

That does not mean rejecting technology. Commercial kitchens are already full of advanced systems, smart equipment, connected monitoring, and increasingly data-led operations. Understanding those technologies, and knowing where they genuinely add value, will become an important part of consultancy knowledge.

And I suspect most consultants will end up somewhere in the middle: using AI where it is genuinely useful. For brainstorming, research, organising information, comparing specifications, speeding up repetitive tasks, or testing early ideas quickly. Not replacing expertise, but supporting parts of the process around it.

The conversation isn’t really about AI. It’s about what good design requires and whether the people paying for it understand the difference between something generated and something considered.

 

Written by Giorgia Lardner