VibeChef

VibeChef turns your fridge into a smart meal planner: scan ingredients, get recipe ideas, and build grocery lists, save time to eat healthier, cut waste, and reduce decision fatigue.

Comment

Inspiration: Welcome to Apartment 2804. Three roommates, three totally different relationships with food, and one shared problem: figuring out what to eat without spiraling into decision fatigue.

When you walk into our kitchen, you'll usually find one of us doing something food-related — rummaging through the fridge, scrolling through recipes, or staring blankly at ingredients we've forgotten to use. We love food, but the daily question of "what to cook?" manages to trip up all three of us in different ways.

There's Julia — our allergy-conscious gourmand. She genuinely enjoys cooking, but her dietary restrictions turn grocery shopping into a puzzle. She wants to be creative, but half the recipes she finds don't fit her needs — too much gluten, too much dairy, too many landmines.

Then there's Fay — the reluctant chef. She'd love to spend less on meal-delivery services like Factor, but cooking feels overwhelming. She doesn't know where to start, what to buy, or how to make meals that won't go to waste.

And then there's me — Claire — the strong cook with an enthusiasm-to-boredom pipeline. I love experimenting with new ingredients, I batch-cook like a machine, and I can make just about anything… but I still get stuck. I forget what's in my fridge. I get bored easily. I buy ingredients that rot before I get inspired. And I waste way too much time deciding what to make.

Living together made something obvious: Even with different cooking levels, dietary needs, and motivations… we're all exhausted by the same mental load.

So I wanted to build something just for us — something that understands how real people cook in real kitchens with real constraints. Not another recipe app. Not another meal planner. But a personal cooking companion that remembers your preferences, evaluates your pantry, and tells you exactly what you can make tonight.

That's how VibeChef was born: a personal, mood-based meal inspiration engine created in Apartment 2804 for three roommates who know food matters — but mental energy matters more.

What it does: VibeChef is a mobile-first cooking assistant that combines

1. Personalized Onboarding

Captures everything that affects cooking decisions:

  • Dietary restrictions
  • Calorie goals
  • Typical cooking time
  • Favorite cuisines & proteins
  • Spice tolerance
  • Effort level preference
  • Batch cooking preference
  • Budget emphasis

2. Pantry Intelligence

Users stock their fridge, freezer, and pantry. The app stores:

  • Ingredient name
  • Quantity
  • Location (fridge/freezer/pantry)
  • Expiration date

3. Recipe Recommendation Engine

Uses a custom scoring formula to rank recipes:

$$\text{Score} = 0.35 \cdot \text{PantryMatch} + 0.25 \cdot \text{VibeMatch} + 0.15 \cdot \text{BudgetFit} + 0.15 \cdot \text{Novelty} + 0.10 \cdot \text{UseSoon}$$

This lets the app serve dishes that are:

  • New but familiar — introducing variety without overwhelming
  • Aligned with the user's cooking "vibe" — mood-driven meal matching
  • Budget-conscious — respecting financial constraints
  • Practical for batch cooking — supporting meal prep lifestyles
  • Optimized to use ingredients before they expire — reducing food waste

4. Batch-Friendly Design

Recipes include notes on:

  • Freezing instructions
  • Storage guidelines
  • Reheating tips
  • Scaling options (×2 or ×3)

5. A Simple, Motivating UX

  • Vibe-picker chips ("Comfort," "Fresh," "Lazy Night," "High Protein")
  • Dish cards showing pantry match percentage and batchability
  • Inventory & preference editing for easy updates
  • One-click shopping lists for missing ingredients
  • Cook history logs to track favorites and avoid repetition

How we built it

We've used Lovable AI to scaffold the full app layout and components.

Architecture:

  • Custom relational schema supporting ingredients, inventories, recipes, tags, preferences, and cooking history
  • Rule-based logic for recipe scoring and filtering
  • JSON-driven recipe steps for easy scaling and editing

Challenges we ran into

1. Modeling Three Very Different User Types

Designing an app that works for a confident cook (Claire), a restricted eater (Julia), and a reluctant cook (Fay) forced us to think deeply about flexibility versus simplicity.

Solution: We built onboarding that's personal but not overwhelming, and structured the data model so dietary restrictions and cooking confidence can meaningfully shape recommendations without adding cognitive load.

2. Reducing Boring Tasks

Pantry entry could easily feel tedious — a major abandonment risk.

Solution: We designed a system with preset common ingredients, quick quantity buttons, and free-form adds to minimize friction while maintaining accuracy.

3. Building a Smart Recommendation Engine Without Overcomplicating It

We needed the logic to feel personal, not random. Balancing multiple factors like pantry match, vibe match, cost, and novelty was a big iterative challenge.

Solution: We weighted the scoring formula based on what actually drives cooking decisions in our apartment, then tested it against real scenarios.

4. Making "Vibe-Based Cooking" Feel Real

Translating moods (comforting, fresh, lazy-night, adventurous) into tags and decision logic required thoughtful design — it's easy to make this feel gimmicky.

Solution: I mapped vibes to concrete recipe attributes (cooking time, cuisine style, ingredient complexity) and validated the mappings through user testing with Julia and Fay.

Accomplishments that we're proud of

Building a fully functioning personalization engine in under 24 hours

We didn't just create a recipe app — we created a system that understands dietary restrictions, cooking time, mood, budget, and batch preferences, and turns all of that into curated meal suggestions.

Designing an onboarding experience that feels human, not clinical

We're especially proud of how the onboarding flow captures real-life cooking behavior without overwhelming users.

Modeling three radically different user types in a single product

VibeChef works for:

  • A strong home cook (Claire)
  • A dietary-restricted cook (Julia)
  • A reluctant cook who relies on meal delivery (Fay)

Building something flexible enough for all three was a real achievement.

Creating a robust data schema that can scale

The relational database handles ingredients, recipes, tags, pantries, preferences, and history without collisions or ambiguity.

Making vibe-based cooking feel real

Turning "comforting" or "adventurous" into actual logic that influences suggestions was complex, and we're proud we pulled it off.

Reducing mental load in a way that actually feels delightful

Our dish cards, vibe chips, and inventory prompts all feel like a warm conversation with a mini sous-chef, not another to-do list.

What we learned

1. Cooking is emotional, not just functional

What people cook isn't just determined by hunger, but rather shaped by mood, time, stress, and constraints. Designing for different users with distinct needs reinforced that personalization is essential.

2. Personal AI = great onboarding + great data

Good data → great suggestions. Bad or missing data → frustration. Onboarding ended up being as important as the actual recommendation algorithms. The quality of the input directly determines the quality of the experience.

3. AI coding tools shine with great prompts

Lovable worked incredibly well once the schema, screens, and flows were fully defined. Clarity in requirements unlocks speed in execution.

4. Small friction points matter

Even 2–3 extra taps can make a user abandon pantry entry or preference setup. Micro-interactions mattered more than expected — streamlining every step was crucial to keeping users engaged.

What's next for VibeChef

We have several extensions in mind that would take VibeChef from "helpful app" to "intelligent kitchen companion."

1. AI-Generated Weekly Meal Plans

Move from single-suggestion recommendations to 3–5 day dynamic meal plans based on:

  • Schedule
  • Leftovers
  • Budget
  • Nutrition goals

2. Automated Pantry Input

Add features like:

  • Barcode scanning
  • Receipt scanning
  • OCR recognition
  • Auto-tracking expiration dates

This would eliminate manual data entry entirely.

3. Smart Shopping Lists

Generate optimized grocery lists based on:

  • Weekly plans
  • Sales at local stores
  • Minimal new ingredients needed
  • Seasonality

4. Leftover Remix Mode

Automatically suggest ways to turn leftovers into new dishes (e.g., roast chicken → grain bowl → tacos).

5. Nutrition Insights

Lightweight tracking of macro balance, fiber, variety, or protein — no calorie obsession.

6. Cross-User Support

Enable profiles for roommates, partners, or families so everyone's needs are considered in the suggestions.

7. Generative Recipe Creation

Use AI to synthesize new recipes from your pantry, preferences, and mood — not just recommend existing ones.

8. Social Features

Let users share:

  • Favorite dishes
  • Batch cooking routines
  • Remix ideas

Built With

  • lovable
Share this project:

Updates