Skip to main content

​
General Questions

​
What is Synheart?

Synheart is a privacy-preserving platform for building human-state-aware applications. It consists of two main SDKs:
  • Synheart Wear: Streams biometric data from wearables
  • Synheart Emotion: Detects emotions from biosignals on-device

​
Is Synheart free to use?

Yes! Synheart SDKs are open-source and free to use. Check individual repository licenses for details.

​
What platforms are supported?

  • Mobile: iOS (Swift), Android (Kotlin), Flutter (Dart)
  • Desktop/Backend: Python
  • CLI: Python-based development tool

​
Is Synheart HIPAA/GDPR compliant?

Synheart SDKs are designed with privacy in mind:
  • All processing happens on-device
  • No data is sent to external servers
  • Users control their data
  • However, compliance depends on how you implement and use the SDKs in your application. Consult with legal counsel for your specific use case.

​
Synheart Wear

​
Which wearables are supported?

Currently supported:
  • ✅ Apple Watch (via HealthKit)
  • ✅ WHOOP (via cloud API)
  • ✅ Health Connect (Android)
  • 🔄 Fitbit (partial support)
  • 🔄 Garmin (partial support)
  • 📋 Samsung Watch (planned)

​
Do I need OAuth credentials for all wearables?

No. Only cloud-based wearables (WHOOP, Garmin, Fitbit) require OAuth credentials. Local wearables (Apple Watch via HealthKit, Health Connect) use system permissions.

​
Can I use multiple wearables at once?

Yes! The SDK supports multiple device adapters simultaneously. Data from all devices is normalized to the same schema.

​
How often does data update?

Update frequency depends on:
  • Device capabilities (some update every second, others every minute)
  • SDK configuration (you can set polling intervals)
  • Network connectivity (for cloud wearables)
Typical ranges: 1-60 seconds for heart rate, 5-60 seconds for HRV.

​
What data formats are supported?

All SDKs output the unified Synheart Data Schema v1.0 (JSON format). You can also export to CSV for analysis.

​
Synheart Emotion

​
What emotions can be detected?

Currently supports three emotions:
  • 😊 Amused: Positive, engaged state
  • 😌 Calm: Relaxed, peaceful state
  • 😰 Stressed: Anxious, tense state

​
How accurate is emotion detection?

The model achieves 78% accuracy on the WESAD dataset for 3-class emotion recognition. Accuracy may vary in real-world conditions.

​
How fast is emotion detection?

Inference latency is < 5ms on modern mid-range devices. The SDK processes data in real-time with minimal overhead.

​
What data do I need for emotion detection?

You need:
  • Heart rate (bpm)
  • RR intervals (milliseconds between heartbeats)
Minimum window: 60 seconds of data recommended for stable results.

​
Can I use emotion detection offline?

Yes! All emotion detection happens entirely on-device. No internet connection required.

​
Is emotion detection a medical device?

No. Synheart Emotion is for wellness and research purposes only. It is not a medical device and should not be used for medical diagnosis or treatment.

​
Integration

​
How do I combine Synheart Wear and Synheart Emotion?

Stream data from Synheart Wear and feed it into Synheart Emotion: 
// Stream from wearable
synheartWear.streamHRV(windowSize: Duration(seconds: 60))
  .listen((metrics) {
    // Feed into emotion engine
    emotionEngine.push(
      hr: metrics.getMetric(MetricType.hr)!,
      rrIntervalsMs: metrics.rrIntervals,
      timestamp: DateTime.now(),
    );
    
    // Get emotion results
    final results = emotionEngine.consumeReady();
    // Use results...
  });

​
Can I use just one SDK?

Yes! You can use Synheart Wear without Synheart Emotion, or use Synheart Emotion with your own biosignal data source.

​
How do I handle permissions?

Each platform has different permission requirements:
  • iOS: HealthKit permissions via Info.plist and requestAuthorization()
  • Android: Health Connect permissions via system settings
  • Cloud APIs: OAuth flow for WHOOP, Garmin, Fitbit
See platform-specific guides for details.

​
Development

​
Can I test without a physical wearable?

Yes! The CLI tool allows you to:
  • Pull historical data from cloud APIs
  • Simulate data streams
  • Test OAuth flows locally

​
How do I debug issues?

  1. Enable debug logging in SDK configuration
  2. Check device logs (Xcode Console, Android Logcat)
  3. Use CLI tool for testing cloud integrations
  4. Review troubleshooting guide

​
Where can I find code examples?

​
Privacy & Security

​
Does Synheart send data to servers?

No. All processing happens on-device. No biometric data is sent to Synheart servers or any external servers.

​
How is data stored?

  • Local storage: Encrypted with AES-256 (if enabled)
  • Memory: Data is processed and discarded (not persisted)
  • You control: What gets stored and where

​
Can users delete their data?

Yes. Users can:
  • Revoke permissions (stops data collection)
  • Clear app data (removes local storage)
  • Uninstall the app

​
Performance

​
Will Synheart drain my battery?

Minimal impact when configured properly:
  • CPU: < 2% during active streaming
  • Memory: < 3 MB footprint
  • Battery: Similar to other health tracking apps
Optimize by adjusting streaming frequency and window sizes.

​
How much storage does Synheart use?

  • SDK size: ~2-5 MB per platform
  • Model files: < 100 KB
  • Runtime memory: < 3 MB

​
Support

​
Where can I get help?

  1. Documentation: Check relevant guides
  2. GitHub Issues: Report bugs or request features
  3. Discussions: Ask questions in community forums
  4. Email: support@synheart.ai

​
How do I report bugs?

Create an issue on GitHub with:
  • SDK version
  • Platform and version
  • Device model
  • Steps to reproduce
  • Error messages/logs

​
Can I contribute?

Yes! Synheart is open-source. Contributions welcome:
  • Code improvements
  • Documentation updates
  • Bug fixes
  • Feature requests
Check individual repositories for contribution guidelines.
Author: Israel Goytom