A major institutional caterer cut menu planning from weeks to a single day using AI — what that actually means for a kitchen's workflow

The number that gets quoted, and the one that matters more

A major Paris-based institutional catering group has deployed an AI menu tool that generates and manages menus from a 400-recipe seasonal pool in a single day — a process that reportedly took 2 to 4 weeks of manual planning before. That’s the number that gets quoted. The more operationally interesting fact is what it’s actually replacing: not the creativity of menu planning, but the repetitive constraint-checking (allergens, nutrition targets, ingredient availability, rotation rules across a seasonal pool) that used to consume most of that 2-4 week window.

Robotics, deployed for a narrower reason than the headlines suggest

Alongside the AI menu tool, the group has around 200 robots in operation across cleaning and delivery tasks, with kitchen robots for repetitive or genuinely high-risk tasks under active evaluation rather than full deployment. The distinction matters: this isn’t “robots replacing chefs” — it’s automation aimed specifically at the lowest-judgment, highest-repetition, or highest-injury-risk parts of a kitchen’s workflow, which is a much narrower and more defensible claim than the general “AI in restaurants” framing usually implies.

Why the speed-up is real, not just a demo number

A 400-recipe seasonal pool, checked for rotation and constraint compliance across thousands of sites, is exactly the kind of combinatorial problem that’s genuinely slow for a human planner and genuinely fast for a properly constrained AI system — not because the AI is “smarter,” but because checking a recipe against a fixed rule set is a mechanical task at that scale, and mechanical tasks are where this generation of AI tooling is actually reliable, as opposed to open-ended creative recipe invention, which is a different and much less proven claim.

What this means for a kitchen considering AI-assisted planning

The realistic value of AI in recipe and menu workflows right now is concentrated in exactly the part of the job this caterer’s numbers describe — constraint-checking at scale — not in replacing a chef’s actual judgment about what a dish should be.

• The time savings show up in constraint-checking, not creativity — allergen cross-checks, nutrition targets, and rotation rules across a large recipe pool are where AI genuinely accelerates the work.
• Multi-site consistency is the real prize, since the same tool checking the same constraints across thousands of sites is what makes a “one recipe, everywhere” standard actually enforceable at scale.
• Recipe standardization has to happen before AI acceleration is useful — an AI tool checking a messy, inconsistent recipe base just moves the mess faster, it doesn’t fix it.

Nooko, CalcMenu’s AI recipe assistant, already covers the fast-generation half of this: type a dish, an ingredient or a craving and get a professional-kitchen recipe on the spot, allergens, nutrition and cost calculated automatically. Pair that with BlazeChef AI standardizing and adapting those recipes across your recipe base, and the two pieces this kind of speed-up depends on — fast generation and constraint-checking — are already built into CalcMenu.

---

Curious what AI-assisted recipe generation and standardization looks like for your own menu? Book a free 15-minute call with our team — no commitment: Schedule a call.

Sources

• How foodservice giant Sodexo is embracing AI and robotics to reshape the kitchen — Fortune

Explore CalcMenu's recipe management software for restaurants, hotels & catering to see how it applies to your kitchen.