Technology

Understanding human state
through vision, voice, and physiology.

Video

Audio

Physiological Signals

Human State Intelligence

One engine for every human signal

Core technologies

Four research areas that turn everyday camera and microphone input into human state intelligence.

rPPG

Remote photoplethysmography reads subtle blood-flow changes in facial skin with a standard RGB camera. From these micro-variations we estimate heart rate, heart-rate variability, and related vitals in real time — no wearables, no contact, clinically meaningful signal quality.

72 BPM

Facial Analysis

Dense facial landmarks, expression dynamics, and identity signals are extracted frame by frame. We track emotion, stress, and fatigue cues over time, building a stable picture of how someone actually feels — beyond a single snapshot.

Voice Analysis

Paralinguistic features — pitch, energy, rhythm, spectral texture — carry emotional and psychological state. Combined with live speech-to-text and language patterns, voice becomes a continuous sensor for affect, tension, and cognitive load.

Multimodal Intelligence

Vision, voice, and physiology each tell part of the story. Our fusion models align them on a shared timeline and resolve them into one coherent state model — the foundation for care intelligence that sees the whole person.

Compliance

We align our data practices and security posture with global standards for privacy, health data protection, and information security.

GDPR

Our data handling follows EU General Data Protection Regulation principles for lawful processing, transparency, and user rights.

HIPAA

We apply HIPAA-aligned safeguards for protected health information across collection, storage, access, and transmission.

ISO 27001

In progress

We are pursuing ISO/IEC 27001 certification to formalize our information security management system across people, processes, and infrastructure.

Get started with XiTST

Reach out about multimodal care intelligence, pilots, and partnerships.

Contact us