New Year’s resolutions are hard. Mine were to learn a bit more about Streamlit, to get deeper into using local AI as a data correlation mechanism, and to get fitter/healthier.
So that’s where Train Track comes in. It’s a small Python app I am writing using FastCGI and Streamlit to analyze my Apple Health data, exported and sent to Train Track daily via Health Auto Export. HAE is pretty cool, it takes everything my phone knows about my health and habits and can send it via MQTT, to Home Assistant, or (most important to me) a REST endpoint.
Second image is a local LLM analyzing the data and giving some basic recommendations and correlations. That one needs a LOT more work.
Note: I am not stupid enough to blast my health data into the Fediverse, that’s example content :)