Skip to content
RunMyKitchen

How It All Fits Together

Think of RunMyKitchen as a personal chef who knows three things: what’s in your fridge, what your family eats, and what’s available at your local store. Every time you ask for meal ideas, those three inputs combine to produce recipes that actually work for your household tonight.

The four layers

Section titled “The four layers”

Your Pantry

Everything you have on hand. Add items by typing, chatting, uploading a receipt photo, or snapping a picture of your fridge. The AI parses it all and keeps a running inventory.

Your Household

Family size, allergies, dietary needs, and taste preferences. Set this up once. The AI filters every suggestion through these constraints so you never have to double-check labels.

Recipe Generation

The AI creates recipes from scratch — it doesn’t pull from a fixed database. It combines your pantry, your household profile, and culinary knowledge to generate meals that fit your exact situation.

Shopping & Delivery

Whatever you’re missing gets added to a smart shopping list. One tap sends it to Instacart at your preferred store. You pay normal grocery prices — nothing extra.

How the AI picks recipes

Section titled “How the AI picks recipes”

When you ask “what should I cook this week?” the AI works through a quick decision chain:

  1. Inventory check — What proteins, vegetables, grains, and staples do you have?
  2. Household filter — Remove anything that conflicts with allergies, dietary needs, or strong dislikes
  3. Variety balance — Avoid repeating the same protein or cuisine two nights in a row
  4. Effort matching — Mix quick weeknight meals with slightly more involved weekend dishes
  5. Gap analysis — Figure out what’s missing and build the shortest possible shopping list

The result: a week of meals where most ingredients come from your kitchen, and the shopping list is short.

It gets better over time

Section titled “It gets better over time”

The AI learns from how you use it. Rate a recipe, skip a suggestion, swap a meal — all of that feeds back into future recommendations.

  • Ratings teach it your taste — spice tolerance, cuisine preferences, complexity level
  • Swaps signal what didn’t land — too long, wrong vibe, not enough variety
  • Pantry history reveals your shopping patterns — it stops suggesting ingredients you never buy

You don’t need to configure anything. Just cook and react. The AI adjusts.

The mental model

Section titled “The mental model”

If this all feels complicated, here’s the simple version:

You tell it what you have. It tells you what to cook. You buy what’s missing.

Everything else — household profiles, learning your taste, smart shopping lists — just makes that loop smoother over time. Start with the basics and let the rest build naturally.

Keep going

Section titled “Keep going”