FAQ
Straight to the point — including about the system's limits, ethics, and bias.
The system and its limits
No. FROID is a tool that supports clinical listening — a coprocessor of biological signals, never an autonomous diagnostic instrument. No index (IPM, IDM, dominant zone) should be used as a standalone conclusion; assessment and the therapeutic decision remain the responsibility of the licensed professional.
Not directly. The underlying acoustic physics (spectral energy via FFT) is real. The specific mapping of zones to emotional themes is a FROID proprietary interpretive framework, inspired by voice biofeedback technology — it is not a finding from peer-reviewed research. Full details on the Science page.
Head-on. Voice and face models have documented demographic bias. FROID's main mitigation is not using absolute tables: every patient is compared only against their own baseline, calibrated in the first 60 seconds of the session, measuring deviation relative to themselves. Even so, this does not eliminate all bias — the Ethics, Fairness & Limits page describes the safeguards and remaining limitations.
The system does not try to guess at corrupted data. Real-time triage layers (Quality Gates) monitor facial visibility and acoustic noise; when quality drops, the dashboard alerts the professional and reduces the weight (reliability factor) of the compromised channel, preventing the indicator from drifting due to network noise.
No. FROID does not automatically detect emergencies or imminent life risk. In the presence of crisis signs, the professional should activate conventional channels — local emergency services and crisis lines, and the protocols of their practice. The system supports perception, not emergency surveillance.
Yes. Audio, video, and transcript capture each require free, informed, and granular consent (separate authorizations), with the right to revoke at any time and permanent data deletion (LGPD article 18 in Brazil). We provide a consent form template aligned with CFP/CFM resolutions on the Ethics page.
Phonation processing and facial muscle imaging happen locally, on the practice's computer (Edge AI). Raw biometric files are not transmitted to public clouds for inference. Details on the Security page.
Audio and video are processed for metric extraction according to the patient's consent as configured by the professional. See the Security page for the applied data controls.
No — FROID runs in the browser via the Web Audio API and a standard webcam, with no need for dedicated sensors.
The first 60 seconds establish the patient's individual baseline (F0, energy per band, resting facial state). Every subsequent reading is compared against this reference, not against a generic population.
Yes, the model is per-session credit packages, with no mandatory monthly fee or minimum commitment. The PLUS and MASTER plans come with guided onboarding — talk to the team at About & Contact.
It is optional and depends on explicit consent. It only triggers when the balance reaches zero, repurchases the same package, keeps the chosen currency, and uses the value validated by Stripe. If payment fails, a new action from you may be required.
Each session completed consumes one credit, regardless of plan. There is no mandatory monthly fee.
Prices are shown in US dollars, and checkout also accepts BRL, EUR, and CNY, with the value validated by Stripe at the time of payment. Auto-recharge always keeps the currency chosen on the first purchase.
No — the values published on the Pricing page reflect this initial phase of controlled availability and will be updated soon. Packages already purchased keep the price paid at the time of purchase.