How it works · in 60 seconds

The more it sees, the smarter it gets.

A camera and microphone capture the session. A wearable extends the picture between sessions. A proprietary ML engine fuses 58 signals into a continuous read of progression, risk and alignment, delivered to the operator's dashboard in real time.

Schematic four-stage flow: capture aperture, fused waveforms, observing eye, dashboard pane — connected left to right by a single thin gold line.

Proprietary ML model · Illuminafy Intelligence Engine · PCT WO 2023/017286 A1

I

Reads

Continuous + in-session

58 observable in-session signals plus wearable HRV, sleep and activity between sessions. Continuous, not episodic.

II

Learns

Model grows with data

Every session, every wearable day, every longitudinal context feeds the model. Patterns emerge across thousands of cycles.

III

Predicts

Before it happens

Trajectory modelling flags deterioration weeks before crisis, not after the event.

IV

Adapts

Infinitely contextual

Cultural · emotional · personality · industry. Reshapes itself to every context.

The dashboard · bi-directional intelligence

We see both sides of the room.

Progress for the patient. Alignment for the therapist. One platform measuring both, diagnostic, not judgmental. The therapist retains clinical authority.

Bi-directional dashboard: Patient view with engagement, trajectory, risk and HRV alongside Therapist view with alignment, pacing and prompt suggestions.

Illustrative · stylised numbers · alignment framing is diagnostic, not judgmental.