Apple Health Assistant

In this guide, you will learn how to set up an Apple Health Assistant using PromptQL and Hasura. This assistant empowers you to analyze your own Apple Health data by interacting in natural language. You’ll import your health metrics from an iPhone into a PostgreSQL database, connect it with PromptQL, and interact with the data.

Here’s a sample conversation:

Check out the GitHub repo.

Pre-requisites

You’ll need the Hasura CLI (authenticated via a Hasura account) and Docker installed on your local machine. Links to these steps are below:

Sign up for a Hasura account (if you haven’t)
Hasura CLI installed and authenticated
Docker installed on your local machine

Export your Apple Health Data

This is a private health assistant and it relies on the Apple Health data of the user. In order to export the Health data from your iPhone, follow the instructions below:

Go to the Health app on your iPhone.
Tap your picture or initials at the top right.
If you don’t see your picture or initials, tap Summary or Browse at the bottom of the screen, then scroll to the top of the screen.
Tap Export All Health Data, then choose a method for sharing your data. If you have a Macbook, share it via AirDrop to get started quickly. Otherwise, use Google Drive to save and download.

Here are the docs on Apple Support around this. Do note that the export can take some time depending on how much data you have. Ensure your phone doesn’t lock during the export to prevent interruption.

Import your Apple Health data into PostgreSQL

Step 1: Clone the project

git clone [email protected]:praveenweb/apple-health-ai-assistant.git
cd apple-health-ai-assistant

Step 2: Copy Apple Health export.zip

Copy the export.zip file obtained from the Apple Health export above into the root directory of the git repo you just cloned.

Step 3: Configure environment variables

cp .env.sample .env

The default Postgres credentials should work. Feel free to change them as necessary.

Step 4: Start the data import using docker compose

docker compose up -d –build

This should run the Python program, which takes in the export.zip that was placed in the root directory, unpacks and parses the XML data, and inserts it into the PostgreSQL database.

Note: Depending on how big your health export is, it should take anywhere between 30 seconds to 2 minutes to fully import the data. The schema will be initialized quickly and the data import happens in the background, so you can proceed to follow the steps below.

Build your PromptQL app

Now, let’s set up the Hasura DDN project with PromptQL to start exploring the data in natural language!

Set up the Hasura DDN project already scaffolded in the repo:

cd ddn-project
ddn supergraph build local
ddn project init

Add Anthropic API key to the.env.

Get an api key from https://console.anthropic.com/settings/keys

...
ANTHROPIC_API_KEY=<your-anthropic-api-key>

To use an OpenAI API key instead, you’ll have to set the OPENAI_API_KEY key-value in your .env file and change the environment variable LLM to openai in the compose.yaml file.

Start the DDN project

Let’s start the DDN project by executing the following command:

ddn run docker-start

Open the local DDN Console to start exploring:

ddn console --local

This should open up your browser (or print a browser URL) for displaying the Hasura Console. It’ll typically be something like: https://console.hasura.io/local?engine=localhost:3280&promptql=localhost:3282

Ask questions about personal health

The app has metadata about various health metrics and it can answer a range of questions; as simple as “What’s my step count per day?” to as complicated as, “Give me a nutrition plan based on my weight / BMR, activity and heart health to reduce 1 pound every 2 weeks.”

Here’s a sample of what you can ask:

What’s my average walking per day?
How much time am I spending outdoors?
Based on my weight and BMR, tell me if I’m on track to lose weight and what should my calorie intake be for losing 1 pound per week. Analyze data over the last 3 months.
Am I being exposed to a lot of noise in the environment? Check the impact on my hearing health.
How is my heart health?
How active have I been over the last few months? Are there any patterns and anomalies in my activity?
How much exercise activity did I do in the last year? What’s the trend looking like?

Clean up and restart your app

If you want to reset the Apple Health data and start from scratch, you can execute the following in the root directory of the repo:

docker compose down -v && docker compose up -d --build

Interactive demo

Here is an interactive demo of the AI Assistant. We’ll start by looking at the exported data and then talk to it with PromptQL! You can click the Replay button at any point to restart the demo.

PromptQL Guide Realtime Chat with BART API