Personal Health Data Integration via AI Assistants
reThrive Labs LLC has launched native iOS and Android applications for its health-data service, freddy, enabling users to bridge the gap between on-device health metrics and AI models. By connecting Apple Health and Android Health Connect, the platform allows users to query fragmented health data—such as heart-rate variability, sleep patterns, and glucose levels—using AI assistants like ChatGPT and Claude through the Model Context Protocol (MCP). According to reThrive Labs founder Tom Tomaszewski, the mobile expansion ensures that data residing on a user’s phone can finally be integrated into the tools used for cognitive tasks.
Bridging the Mobile Health Data Gap
Before the release of these mobile apps, web-based health services faced a significant technical constraint: they could not access on-device health repositories. Apple Health and Google’s Health Connect aggregate data from wearables and third-party apps, but this information remains siloed within the phone’s operating system. The freddy mobile app runs in the background to sync this local data, making it queryable alongside cloud-based sources like Oura, WHOOP, and Dexcom. This allows for a unified view of health metrics that previously required switching between multiple specialized dashboards.
The Model Context Protocol (MCP) acts as a universal bridge, allowing AI models to securely access external data sources. By using MCP, freddy allows users to keep their health data private while still being able to ask natural language questions about their long-term recovery trends or workout performance.
Privacy and Data Sovereignty
The service operates on a direct-to-consumer model where the user maintains control over their data permissions. According to the company, freddy provides a read-only interface that does not train AI models on the personal health information it processes. Users connect their accounts once to generate a private URL, which is then used to grant specific AI clients access to their data. This architecture aims to solve the problem of “disconnected apps” that currently prevent users from seeing a holistic view of their personal health metrics.
Future Trends in AI-Driven Personal Health
The integration of personal health data with Large Language Models (LLMs) points toward a shift in how individuals interact with their biometric metrics. Rather than manually interpreting complex charts, users are moving toward conversational health insights.
If you use multiple health platforms like Garmin for workouts and Oura for sleep, try consolidating them into a single AI-ready pipeline. This reduces the cognitive load of checking five different apps each morning to understand your readiness score.
Frequently Asked Questions
What devices does freddy support?
freddy supports a wide range of devices and platforms, including Oura, WHOOP, Garmin, Polar, Withings, Dexcom, Suunto, Hevy, Concept2, and Intervals.icu, as well as native integration with Apple Health and Android Health Connect.
Does freddy train AI models on my data?
No. According to reThrive Labs, the service is read-only, and it does not train models on user data.
Is the freddy service available now?
Yes, the native applications are currently available for download on the Apple App Store and the Google Play Store.
Have you struggled to aggregate your health data across multiple wearables? Share your experience in the comments below or subscribe to our newsletter for more updates on the future of personal health technology.
Keep reading
