Open Wearables

What is Open Wearables?

Open Wearables is an open-source, self-hosted wearable API and health intelligence platform. It connects wearable and health-tracking data sources and turns them into health scores and AI reasoning frameworks that can generate recommendations based on trends and anomalies.

The platform is designed for teams that build health products or dashboards. It provides a unified, normalized API for ingesting wearable data, open health scoring algorithms, and a structured “health AI engine” that produces audit-ready reasoning instead of returning raw metrics.

Key Features

• Unified wearable data ingestion via one API: Connects wearables and health sources such as Apple Health, Whoop, Oura, and Samsung Health, with normalized and deduplicated data handled by the platform.
• Self-hosted deployment on your infrastructure: Runs within your environment so data ingestion and scoring occur under your control.
• Open health scoring algorithms: Provides open algorithms for sleep, recovery, strain, stress, HRV, VO2 max, and related metrics, with the ability to audit and tune thresholds.
• Health AI engine with structured reasoning: Detects trends and flags anomalies across scores, producing recommendations tied to patterns rather than reading off standalone numbers. Includes an MCP server for connecting to an LLM.
• Configurable coaching profiles by domain: Lets you define how the engine reasons for different use cases (e.g., wellness, performance, clinical monitoring) while keeping scores consistent across devices.

How to Use Open Wearables

1. Start with the platform setup (via the site’s getting started flow) and deploy it in a self-hosted environment.
2. Connect a wearable/health data source (for example Apple Health, Whoop, Oura, or Samsung Health) through the platform’s wearable API.
3. Use the scoring layer to compute open health scores (e.g., sleep quality, recovery, strain, stress, HRV-related measures) and tune thresholds for your population.
4. Run structured reasoning using the health AI engine to identify trends and anomalies across scores, then generate recommendations appropriate to your coaching profile.

Use Cases

• AI coaching for fitness and recovery: A product team builds a coaching feature that combines scores (such as strain, recovery, and sleep) to recommend actions like reducing intensity or prioritizing sleep based on multi-day trends.
• Longevity and long-term health optimization: Developers create protocols and dashboards that track aging- and wellness-adjacent biomarkers or long-term trends derived from users’ wearable data, using open scoring and configurable reasoning.
• Corporate wellness monitoring: An organization deploys self-hosted scoring and reasoning to generate sleep, stress, and recovery insights across a group while keeping data on its infrastructure.
• Clinical monitoring with auditability: A clinical or healthcare-adjacent team uses open algorithms so clinical staff can verify the components behind health scores and the reasoning framework.
• Personal health dashboard experiences: Teams build applications that surface consistent health scores and recommendations to end users, regardless of which supported wearable or device they use.

FAQ

• Is Open Wearables a wrapper around an LLM? The platform describes its health AI engine as a structured health reasoning framework (with an MCP server for LLM integration), not “a wrapper.”

• Can the scoring and reasoning be audited or customized? Yes. The site states that the health scoring algorithms are open (audit-ready) and that thresholds can be tuned; coaching profiles define the engine’s reasoning for different domains.

• Can I run it without sending data to a third party? The platform is self-hosted on your infrastructure, and the site emphasizes patient data never leaving the premises in the clinical monitoring use case.

• What devices and health sources does it support? The page lists integrations including Whoop, Garmin, Oura, Apple Health, Strava, Polar, Suunto, Samsung Health, Google Health Connect, Ultrahuman, Fitbit (and mentions Coros and Xiaomi as “soon”).

• Does it provide health scores across multiple wearables? The site describes unified scoring so users receive the same scores regardless of which supported wearable they use.

Alternatives

• Proprietary wearable analytics APIs: Instead of an open, self-hosted stack, these typically provide black-box scoring and closed logic delivered as hosted APIs. They may be faster to start but offer less auditability and tuning control.
• In-house pipelines plus custom scoring: Teams can build their own data ingestion and scoring logic. This can match specific requirements but shifts the work of normalizing data, implementing scoring algorithms, and maintaining updates to your team.
• General LLM + metrics dashboards: Using LLMs to summarize raw wearable metrics can produce narrative output, but it does not provide the platform’s structured reasoning framework, open scoring algorithms, or unified coaching profiles.
• Health data interoperability tools: Alternatives may focus on device data synchronization (moving data into a central store) without providing the scoring and reasoning layers described by Open Wearables.