mindfull

Welcome Screen
Cookbook - Saved Recipes
Cookbook - Selected Recipes
Grocery List
Medication List
Adding a New Medication
Camera Pill Detection
Added Medication to List

Inspiration

We built Mindfull for ourselves and friends who struggle with executive function, especially ADHD. We wanted a dead-simple, mobile first assistant that removes decision fatigue and adds gentle, real accountability.

What it does

mindfull is a life assistant app built specifically for neurodivergent needs. In just one place, it takes the stress out of daily routines with two powerful features. First, our Meal Planner: you can input or generate meals, and our AI instantly creates a precise grocery list tailored to the number of people you’re cooking for — no overbuying, no mental math. Second, our Medication Reminder: it doesn’t just ping you, it makes sure meds actually get taken. The app requires a quick photo confirmation before the alarm turns off, adding accountability while staying simple to use.

How we built it

React Native (mobile app): Builds the UI for Meal Planner and Med Reminders, handles input and navigation, and triggers actions like “generate recipe” or “confirm medication taken.”

Node.js + Express (API): Exposes routes, validates requests, calls the Gemini helper, and orchestrates the flow from prompt → instructions/ingredients → database writes.

Supabase (Postgres + Auth): Stores users, recipes, and ingredients. After creating a recipe row, we bulk-insert its ingredients with the same recipe_id; Auth provides the user_id (UUID) and RLS policies.

Google Gemini (AI): Generates concise cooking steps and a structured ingredients list (JSON). We parse/repair the JSON, sanitize it, and persist the results in the linked recipes and ingredients tables.

Challenges we ran into

Wrangling SDK/auth (ESM vs. CommonJS, API keys) and enforcing structured JSON from the model without breaking the flow. Supabase RLS and UUID handling, plus getting reliable mobile permissions/UX for camera-based confirmations.

Accomplishments that we're proud of

An end-to-end “idea → recipe → grocery list” flow that saves to the database with zero manual formatting. A medication flow that adds accountability without being punitive, and a schema that cleanly links recipes to their ingredients.

What we learned

Supabase (Postgres + Auth + Storage): New to us—handled accounts and stored recipes/ingredients with row-level security. We also use it to persist the AI output and (optionally) stash photo confirmations via Storage with recipe/med logs linked by IDs.

Expo (managed workflow): Our first time using Expo for mobile—hot reload, painless device testing with Expo Go, and simple builds made iteration on Meal Planner and Med Reminders fast.

Camera capture (Expo Camera / Image Picker): Implemented the “photo confirmation” for medication—request permission, snap a quick pic, timestamp it, and associate it with the reminder/record so the alarm can’t be dismissed without proof.

First-time Mobile App (React Native): Learned mobile UI patterns, navigation, and platform permissions from scratch, wiring screens to the API and Supabase so flows feel native and low-friction.