Shen.AI SDK Documentation

Shen.AI SDK (Software Development Kit) by MX Labs (opens in a new tab) is a B2B platform with precise and easy-to-use camera-based diagnostics of vital signs and wellness.

Evaluating the SDK

You can try Shen.AI SDK on your own device or in the web browser using our publicly available demos - see (opens in a new tab)

What metrics are currently available?

Our SDK currently exposes the following metrics based on real-time video analysis:

  • Heart Rate (real-time and aggregated)
  • Interbeat Intervals (real-time and aggregated)
  • Heart Rate Variability (HRV) - SDNN, lnRMSSD
  • Breathing Rate
  • Cardiac Stress
  • Blood Pressure (systolic and diastolic) - beta version

See Results for more details on the video-based metrics.

The SDK also computes Cardiovascular health risks based on provided risk factors.

What platforms are supported?

The current version of the SDK supports mobile and web platforms.

The following native SDKs are available:

  • Android: Java/Kotlin
  • iOS: Swift/Objective-C
  • Web (Desktop and Mobile browsers): JavaScript/TypeScript

The following cross-platform frameworks are supported (for iOS/Android):

  • Flutter
  • React Native

See system requirements for more details.

We are open to expanding our platform support based on customer demand - feel free to contact us at (opens in a new tab)

How does it work?

The SDK connects to the camera of a mobile/desktop device and uses real-time Computer Vision to isolate a human face in the video stream. A 3d model of the face is constructed, tracked and used to extract a high-quality, dense signal of blood pulsations using remote photoplethysmography (rPPG). Based on that dense signal, the SDK accurately determines the timings and shapes of all observed heartbeats and provides the user with precise heart-related metrics. All computations happen locally on the end device / browser.

What does it consist of?

The SDK consists of:

  • compiled machine code (native shared library on Android/iOS, webassembly on Web) that contains high-performance real-time Computer Vision algorithms and neural networks
  • embedded UI guiding the user through the measurement process and presenting the results
  • platform/framework-specific components for ease of integration and camera access
  • example app code (opens in a new tab) for each supported platform