I exported my data and there was 989MB of data! So I needed to write some code to break this down. The code takes in your export data and gives you options to look at Steps, Distance, Heart rate, Sleep and more. It gave me some cool charts.
I was really stressed at work last 2 years. I decided to pass this data to A.I. It gave me some CRAZY insights:
Seasonal Anomalies: While there's a general trend of higher activity in spring/summer, some of your most active periods occurred during winter months, particularly in December and January of recent years.
Reversed Weekend Pattern: Unlike most people who are more active on weekends, your data shows consistently lower step counts on weekends, suggesting your physical activity is more tied to workdays than leisure time.
COVID Impact: There's a clear signature of the pandemic in your data, with more erratic step patterns and changed workout routines during 2020-2021, followed by a distinct recovery pattern in late 2021.
Morning Consistency: Your most successful workout periods consistently occur in morning hours, with these sessions showing better heart rate performance compared to other times.
You can run this on your own computer. No one can access your data. For the A.I. part, you need to send it to chatGPT or if you want privacy use your own self hosted LLM.
It's also got a docker image and a branch for a still in progress API. This is my first opensource project so any help would be appreciated!