Introduction
Sensia API is a human-centered prediction API that infers cognitive and emotional performance from biometric and sleep data. By analyzing heart rate, respiratory rate, and sleep data, Sensia API returns real-time scores for attention, memory, reasoning, stress, and emotional state.
What you can build
- Wellness apps: provide users with insights of their cognitive and emotional state and actionable recommendations to improve them
- Wearable Companions apps: integrate the Sensia API to stream live mental and physiological state insights directly from wearables
- Sports performance systems: Use the Sensia API to monitor athlete mental load and optimize training, recovery, and performance decisions in real time
- Driver monitoring solutions: Use the Sensia API to assess driver alertness and fatigue levels in real time
How it works
At its core, Sensia API takes a single biometric snapshot, composed of real-time physiological data such as heart rate and respiratory rate as well as sleep data, and runs it through a proprietary model.
Optionally, you can supply up to five historical sessions alongside their previously returned scores to provide temporal context and improve result accuracy.
┌───────────────────────────────────────────┐
│ Input: current_session │
│ (HR, RR, sleep architecture, biometrics) │
└────────────────────┬──────────────────────┘
│
┌───────────┴───────────┐
│ Optional: history │
│ past_sessions[0..4] │
│ past_targets[0..4] │
└───────────┬───────────┘
│
▼
┌───────────────────────┐
│ Feature extraction │
│ + temporal encoding │
└───────────┬───────────┘
│
▼
┌───────────────────────┐
│ Prediction model │
└───────────┬───────────┘
│
▼
┌───────────────────────┐
│ PredictResponse │
│ C_SCORE, stress, ... │
└───────────────────────┘
Output metrics
| Score | Range | Description |
|---|---|---|
C_SCORE | 0–100 | Composite cognitive performance |
attention | 0–100 | Sustained focus capacity |
memory | 0–100 | Working memory performance |
reasoning | 0–100 | Logical reasoning ability |
stress | 0–100 | Predicted stress level |
valence | "Low | High" |
arousal | "Low | High" |
All scores are on a 0–100 scale. Higher is better for all metrics except stress.
For more information on what each score means, see the Output Scores documentation →.