Start from what you have

Everyone has experienced it: staring into a fridge full of random ingredients with no idea what to make. Existing recipe apps assume you'll shop for a specific meal. CookPal works the other way. You tell it what you have and it figures out what you can cook.

A friend, not a firehose

Early AI cooking tools dump 50 recipes on you and call it helpful. We wanted CookPal to feel like a knowledgeable friend in the kitchen, one who understands your constraints, respects your skill level, and gives you just enough options. Not so many that you give up.

Decision fatigue kills motivation

We interviewed home cooks across skill levels, from college students surviving on instant noodles to parents meal-prepping for a family. The pattern was consistent: people don't fail at cooking because they lack skill. They fail at deciding what to cook. Decision fatigue kills motivation before the stove even turns on.

Wet hands, greasy screens

People's hands are wet, greasy, or covered in flour while cooking. Swiping through recipe steps on a phone is a terrible experience. Any cooking tool needs to work hands-free, or at minimum, require only the simplest possible interactions.

Different moments, different speeds

We designed three ways to tell CookPal what you have: type it manually with smart autocomplete, scan your fridge or pantry with the camera using ingredient recognition AI, or maintain a persistent inventory that auto-depletes as you cook. Quick scan for tonight's dinner, curated inventory for weekly meal planning.

Three curated picks, not fifty

CookPal shows you three recipe suggestions, not fifty. Each one is ranked by how well it fits what you have, how long it takes, and how difficult it is. Each suggestion shows which ingredients you have and which you're missing. Missing one thing? CookPal suggests a substitute instead of dropping the recipe.

Set once, remembered forever

Real kitchens are full of edge cases. Allergies, fasting, 15-minute time limits. You set your restrictions once and CookPal applies them to every suggestion automatically. The AI also learns your preferences over time. Cuisines you skip stop showing up. Styles you save show up more.

One step at a time

The step-by-step guided mode was designed for messy hands. Large text, high-contrast visuals, voice commands to move between steps, and timers that run on their own. Each step shows only what you need right now. No scrolling, no hunting for where you were.

Warm, calm, and never condescending

CookPal's design language is warm, approachable, and slightly playful. Like a friend who's great at cooking but never makes you feel dumb for asking. Soft gradients, rounded elements, and generous whitespace keep the interface calm even mid-recipe. The color palette uses warm oranges and creams that feel like a comfortable kitchen.

AI meets kitchen usability

CookPal connected AI capability with actual kitchen usability. The inventory-first approach landed well with users who were tired of recipe apps that assumed a fully stocked kitchen. In usability sessions, participants completed recipes without touching their phone after starting the guided mode.

Design lead and product thinker

As one of three team members, I led the UX strategy and visual design end-to-end. I defined the three-input inventory model (type, scan, persistent pantry), designed the hands-free cooking flow from scratch, and made the call to cap suggestions at three instead of surfacing everything the AI could generate. That last decision was the most argued about in the team — and the one the judges praised most.