What would otherwise be lost

The knowledge of a kitchen cannot stay with one person. For most of my career, what made a restaurant good lived in the head chef’s head. If he left, or fell ill, or the sous chef was away, the standard went with him. The owner is held to ransom by the chef. It is the biggest pain in this business, and I have lived it as an owner and as a chef.

Writing recipes down was never the answer. A recipe tells you what to do once. What a cook needs is the logic by which the recipe was made: why this ingredient, why this method, what to change when the season or the supplier changes. Most chefs will not share that, because it is what makes their restaurant different from yours. I decided to share mine, and to build something that could hold it.

ChefPro.ai carries a menu from first idea to final plate cost as one guided workflow. It is not recipe storage or costing software, which keeps and prices the dishes you already have. It develops new menus to a fixed standard, and checks them before they pass.

• One sequence, concept to cost. The standard is applied at every step, not added at the end.
• Fixed rules. Stock bases, allergens and pairing logic are held as rules and applied the same way every time, so the standard does not drift between dishes or between sites.
• A final check. Anything the system cannot support, such as an unverified allergen, is stopped and held for review rather than presented as fact.
• Credit where it is due. Every traditional technique it borrows is credited to its source.

If an ingredient is out of season, or a supplier cannot get it, it shows the closest swaps, who has them and what they cost. A restaurant without a sommelier can have pairing notes to work from. But the authority stays with the people in the kitchen: each answer has to be accepted before the work moves on. At no point is the AI doing your work for you.

For me it cuts research time by 80 per cent. It can work through complex permutations, vegan or halal and the rest, and give me options that I then piece together to my own taste. The AI does not give you the recipe. The chef still creates it.

What it gives back is time. As an owner and chef, I could only do creative work late at night, after service, when everyone had gone home. That matters even more to a small restaurant run by its owner than to a large group. And the knowledge does not stop at the kitchen door: it can go to the people who serve the table too.

My menu for Serumpun Sarawak was created with it: a conservation-driven menu, built on the biodiversity of Borneo, for the new cuisine we showed at Expo 2025 Osaka.

I also teach with it. At Camp of Leaders 2026, on 8 and 9 December at Taylor’s Culinary Institute, the students run a full simulation of a fine dining restaurant through ChefPro.ai, and learn sustainability, conservation and quality assurance as daily disciplines of the kitchen. ChefPro.ai serves culinary schools, restaurant groups, hotels and chefs who want to scale a signature without diluting it.

ChefPro.ai has three founding partners, and it needs all three.

James Won. I bring the culinary standard the platform holds.

Peter Kua. Peter is a data science strategist and co-founder of Growthpro.Asia. He was Head of Data Science and Analytics at REV Media Group, formerly Media Prima Digital, and helped develop Malaysia’s first National Big Data Analytics Framework under MDEC. At ChefPro.ai he owns the verification at the centre of the product: the grounding, the fixed rules, and the check that stops what the evidence cannot support.

Alvin Koay. Alvin is a serial entrepreneur and the founder of Growth.pro, an AI visibility consultancy. At ChefPro.ai he turns the culinary standard and the data discipline into a product that reaches kitchens, schools and food businesses.

In October 2026 Peter and I talked with Elaine Boey on BFM 89.9’s Enterprise Explores about when AI belongs in a commercial kitchen. Listen on BFM 89.9, or read what I said on air.