Skip to content
RunMyKitchen

How the AI Picks Recipes

The simplest way to think about it: RunMyKitchen is a personal chef who knows your fridge, your family, and your taste — and gets better at the job every week.

Here’s what goes into every suggestion.

It starts with your pantry

Section titled “It starts with your pantry”

The AI knows what you have on hand. When you ask “What should I make tonight?”, it doesn’t search a generic recipe database — it builds meals around your actual ingredients. Recipes that use more of what you already have rank higher. Recipes that require a special trip to the store rank lower.

This is why keeping your pantry reasonably up to date matters. The AI can only work with what it knows about.

Your family shapes every suggestion

Section titled “Your family shapes every suggestion”

If you’ve set up your household — allergies, dietary restrictions, preferences, picky eaters — the AI applies all of that before it suggests anything. It’s not filtering after the fact. It’s building recipes that fit your family from the start.

A few examples of what this looks like:

  • Someone is lactose intolerant? No cream-based sauces unless there’s a substitute
  • A kid who won’t eat mushrooms? They won’t show up in weeknight dinners
  • The household is vegetarian on weekdays? Monday through Friday suggestions skip the meat

You don’t repeat these constraints every time you chat. Set them once, and they’re always active.

Ratings teach it your taste

Section titled “Ratings teach it your taste”

Every time you rate a recipe, the AI learns something. Not just “they liked this dish” — it picks up on patterns across your ratings:

  • You consistently rate Thai food higher than Italian? More Thai suggestions
  • Quick meals get 5 stars but Sunday projects get 3? It leans toward speed on weeknights
  • Your family loves bold spice but dislikes sour flavors? The AI adjusts seasoning profiles

This is why a dozen honest ratings are worth more than a hundred saved recipes. Ratings are signal. Saves are bookmarks.

It knows what to use first

Section titled “It knows what to use first”

The AI pays attention to what’s about to expire or what you bought recently. It quietly prioritizes ingredients that need to be used soon, so you waste less food without having to think about it.

You might notice: the day after you add fresh herbs or ripe avocados, they start showing up in suggestions. That’s not a coincidence.

It gets better over time

Section titled “It gets better over time”

In the first week, the AI is working mostly from your pantry and your household settings. It gives solid suggestions, but they’re general.

By week three or four — after you’ve rated some meals, chatted about what you liked and didn’t, and cooked from a few suggestions — it starts to feel like it actually knows you. The suggestions get more specific, more confident, and more likely to land.

This isn’t a fixed algorithm. It’s a feedback loop: you cook, you rate, it learns, it suggests better, you cook again. The more you use it, the less you have to think about what’s for dinner.

Next steps

Section titled “Next steps”