Swiftulin

Swiftulin is a high-precision, privacy-first insulin dosing assistant for diabetics. It uses high-accuracy natural language processing to extract nutritional data from voice commands, providing real-time insulin calculations without storing sensitive medical history.

Core Philosophical Principles

1. Privacy Above All: No meal or insulin logs are persisted. The app is a calculation tool, not a tracker.
2. Clinical Precision: Supports decimal (1.5) and fractional (3/4) portions with 0.1 step adjustments.
3. Safety First: Transparent math displays every step of the dose calculation (ICR + ISF).

Technical Blueprint

• Voice Engine: expo-speech-recognition + custom NLP extractor with fractional support.
• Nutritional Data: USDA FoodData Central API with robust local fallbacks.
• Architecture: Zustand for flow state, React Query for API data, and SQLite (via Drizzle) for clinical settings.

Getting Started (Step-by-Step for Beginners)

This section provides a foolproof, step-by-step guide to cloning, setting up, and running this project on a brand-new computer. While the app is pre-configured to work out of the box, you can optionally use your own USDA FoodData API key for higher rate limits.

Step 1: Prerequisites

Before starting, ensure you have the following installed on your computer:

• Node.js (version 18 or higher): Download and install from nodejs.org.
• Git: Download and install from git-scm.com.
• Expo Go App (Optional): Download "Expo Go" on your physical iPhone or Android to run the app directly on your device.

Step 2: Clone the Repository

Open a terminal (Mac: Terminal, Windows: Command Prompt or PowerShell) and run the following command to download the code:

git clone https://github.com/gmg5651/DS440Project.git

(Replace the URL above if your repository URL differs).

To enter the project folder, run:

cd DS440Project

Step 3: Install Dependencies

Inside the project folder, install all required packages by running:

npm install

(This might take a minute or two depending on your internet connection).

Step 3.5: Configure API Key (Optional)

If you would like to use your own USDA FoodData API key for production use or higher rate limits:

1. Obtain an API key from the USDA FoodData Central API Guide.
2. Create a .env file from the example template:

   cp .env.example .env

3. Open the newly created .env file and replace DEMO_KEY with your actual API key:

   EXPO_PUBLIC_USDA_API_KEY=your_key_here
   

Step 4: Run the Application

Start the Expo development server:

npx expo start

This will display a QR code in your terminal.

📱 Device Support Note: The native iOS and Android versions are currently in active development. For the most stable experience and for all demonstrations/grading, please use the Web version.

• Web (Primary MVP/Demo): Press the w key on your keyboard. Expo will instantly open a new browser tab with your application running!
• Physical Device (In Progress): You can preview the experimental mobile app by opening the Camera app on your iPhone (or Expo Go on Android) and scanning the QR code, but note that some native features may require a custom build and aren't fully stable yet.
• Simulator (In Progress): Press i to open the app in the iOS Simulator (Mac only) or a to open it in an Android Emulator.

Step 5: Run Tests (Verification)

If you need to verify the code runs perfectly without errors, open a new terminal window inside the project folder and run:

npm test