Atum Platform

What Atum provides

Structured observations relative to a personal baseline.

respiratory state

{
  "baseline": "stable",
  "shift": "moderate",
  "persistence": "3 nights",
  "recovery": "incomplete"
}

Observation first. Interpretation remains external.

Each state is grounded in respiratory metrics — the standard measures people rely on, and the advanced signals Atum derives.

respiratory metrics

{
  // standard
  "respiratory_rate": "14.2 /min",
  "rate_variability": "0.8 /min",
  "inhale_exhale_ratio": "1 : 1.6",
  "breath_depth": "normal",
  "regularity": "0.92",

  // advanced
  "baseline_deviation": "+1.3σ",
  "respiratory_stability": "0.78",
  "autonomic_tone": "elevated",
  "recovery_trajectory": "improving",
  "confidence": "0.94"
}

The API exposes a continuous respiratory signal — derived from breathing, structured over time. It is designed to be read programmatically, integrated into existing systems, and queried as a live stream or historical record.

A continuous signal changes what can be built. It allows systems to respond to physiological state — not to snapshots, surveys, or intermittent readings.

Atum observes.
Partners interpret.
Agents act.

Atum does not decide what a change means. It makes change observable.

Partners combine respiratory observations with symptoms, wearables, labs, medical history, behavior and other signals.

Built for AI health systems

• Health Agents
• Sleep Platforms
• Fitness Systems
• Longevity Platforms
• Research Infrastructure
• Clinical Workflows

The platform