Frequently asked questions

Yes. Speech recognition and AI processing run entirely on your computer. No data goes through the internet. You can even disconnect from the network during use — Audioris keeps working normally.

Windows 10/11 or macOS 12+ with 16 GB RAM and 10 GB available disk space. The local AI and speech recognition run simultaneously on your machine: below 16 GB we do not support use. Processor: Apple Silicon, or a recent x64 processor with at least six physical cores (Intel Core i5 12th generation or newer, AMD Ryzen 5 5000 series or newer). Note: some chips sold as “Ryzen 5” have only four cores — the report is then generated noticeably more slowly. Windows machines on ARM are not supported.

Your data stays on your machine. The application stores everything exclusively locally, in a SQLite database on your computer: no data is sent outside, neither to Audioris nor to any third party. At-rest encryption of the database is integrated — AES-256 (SQLCipher scheme), with a key derived via scrypt from a passphrase only you know — and integration testing is under way. It is not yet in service on practitioners' machines. In the meantime, we recommend enabling your operating system's disk encryption (FileVault on macOS, BitLocker on Windows).

Audioris Psy is designed for any professional whose practice relies on attentive listening and structured report-writing: neuropsychologists, clinical psychologists, psychiatrists, psychotherapists, doctors, speech therapists, occupational therapists. Other versions of Audioris are in development for additional demanding sectors.

Yes. The Audioris listening device is part of the initial setup fee. Audioris is designed as a complete software + hardware solution, and the device is calibrated for the application — ensuring capture quality and consistency of experience.

The device has been designed and calibrated specifically for Audioris: 360° pickup covering both clinician and patient, active suppression of ambient noise (keyboard, ventilation, echoes), and a timestamped SHA-256 fingerprint of every recording, which makes any later modification of the file traceable.

The Audioris listening device, software installation on your computer, customization for your practice (headers, profiles, anamnesis templates), and a guided training session. You're operational in about 1 hour.

Audioris works in two phases: an initial setup fee covering the listening device and installation, then a monthly subscription covering software use, updates, technical support during business hours, and free hardware replacement for the entire duration of your subscription. Detailed pricing on request to contact@audioris.eu.

Yes, access is entirely free during the beta period, with no commitment. We notify you well in advance before any move to a paid plan, and you remain free to continue or not.

The Audioris listening device remains yours (you paid for it in the setup fee). Your case files and reports stay on your machine, in their exported formats: they belong to you and are never erased remotely. We commit contractually that ending the subscription never deprives you of data already produced. Recording and generating new reports, however, require an active subscription.

It is rarely asked in advance, and it becomes urgent on the day: the records remain covered by professional secrecy, but no one can reach them any more. Belgium already has an instrument for this: IZIMI, the free digital vault of the Belgian notariat. The practitioner files an instruction there for their notary — what to do with the records, who to entrust them to, or to destroy them. During their lifetime they remain the sole custodian: nothing is shared. We document the procedure and explain it at installation. It is a notarial step, not a software feature.

We do not claim a compliance label: GDPR compliance is demonstrated, not declared, and you remain the data controller for your patients' data. What we can state concerns the architecture: Audioris does not collect, process or transmit any patient data to external servers; transcription, analysis and storage happen entirely on your machine. That is a favourable basis for your own compliance, in particular on data minimisation and the absence of transfers. The security measures still to be delivered — first among them at-rest encryption — are stated above, without evasion. What we no longer say →

Yes, in accordance with the deontological rules applicable to your profession. Collecting consent currently follows your own usual procedure: Audioris does not handle it yet. An informed consent module — digital patient signature at the start of the session, archived with the case file — is on our roadmap, with no announced delivery date at this stage.

Let us be precise about what actually protects you, and about what is still missing. What is in place: the application runs offline and exposes no network service, which greatly reduces the attack surface; it sends no data outside; every recording carries a timestamped SHA-256 fingerprint that makes any modification detectable. What is not in place yet: at-rest encryption of the database and encrypted backups are under development. Until they ship, protecting your case files against theft of the machine relies on your operating system's disk encryption (FileVault, BitLocker) — which we recommend enabling. No vendor can promise you zero risk; we would rather tell you where we stand.

The eID plays two roles: (1) security and traceability — it reliably authenticates the patient sitting in front of you, ensuring the file you open matches the right person; (2) instant pre-fill — surname, first name, date of birth, gender, address and national number are injected directly into the report header. No more manual entry, no more transcription errors, no more duplicate files. Reading takes a few seconds and everything works fully locally.

The eID is a convenience, never a requirement. If the patient forgot it, the card is faulty, the patient is a minor, or has a foreign profile without a Belgian eID, Audioris falls back to classic manual entry: a structured form with the same fields (surname, first name, date of birth, etc.) that you fill in within seconds. The case file is created in exactly the same way, and the patient can perfectly well bring their card to the next session to complete the record. No session is ever blocked by an eID issue.

Today, you tell it: a button lets you tag in one click, during the interview, who is currently speaking — practitioner or patient. It is simple, immediate, and never wrong, because you are the one deciding. Assisted diarization (automatic speaker recognition, with uncertain passages flagged for review) is on our roadmap and is not available yet.

Yes — and it's actually a mandatory step of the Audioris workflow. Our solution produces a structured draft that you review, correct and validate. You sign what you have approved, never what the machine wrote.

Approximately 1 hour for the complete setup: software installation, listening device pairing, practice configuration and training session. You leave operational.

Audioris hardware is guaranteed for the entire duration of your subscription — including beyond the 2-year legal warranty period. In case of failure, we commit to replacement within 72 hours anywhere in Belgium, by direct delivery handled by us. Our goal: zero interruption to your clinical activity.

Yes. Any interested practitioner can request a free personalized demo (30 min via video call) and, if selected among the 20 Founding Members, join the free beta in their practice, with a loaned listening device. No commitment at the end of the beta. Request a demo.