About — Audioris

Our vision

AI serving the professional, not the other way around

Audioris was born from a simple observation: every professional whose work relies on attentive listening — neuropsychologists, psychologists, doctors, lawyers, investigators, coaches, social workers — spends considerable time on administrative writing. This time could be spent on what matters: listening, analyzing, supporting.

AI alone is not enough. It's its combination with your expertise, your reflection and your analysis that transforms your practice. Audioris is the tool serving this synergy, never the other way around — and all your sensitive data remains under your total control, locally, with no cloud connection.

Clinician using Audioris during a consultation

Our values

What guides us

🔒

Privacy first

Your patients' data never leaves your machine. It's a founding principle, not an option.

⚡

Simplicity

A tool that integrates seamlessly into your practice. Install, launch, work.

🤝

Serving the profession

Every variant of Audioris is co-designed with field professionals to meet their real needs.

🇪🇺

Sovereignty

Developed in Belgium. Architecture designed so that your data stays under your sole control, on your machine.

What is proprietary

Five building blocks that are ours

The speech recognition and language models are open building blocks, and we say so. Our asset is the domain layer that turns them into a specialised clinical tool.

1

The clinical domain model

Eleven headings, nineteen anamnesis domains. Our ontology, written with a practitioner. No product surveyed structures its output according to the headings of the psychological anamnesis.

2

Contextual reformulation

The software does not return what was said: it reformulates the verbatim into the register of the clinical record and files it under the heading it belongs to. A transcript returns words; we return a usable document.

3

The chain runs on the machine

Capture, transcription, structuring, generation: everything runs on the practitioner's computer. Cut the network and the product still works. That is the operational definition of sovereign AI.

4

Anchored in Belgian infrastructure

Reading of the national identity card, French-speaking Belgian clinical terminology, compliance with professional secrecy under article 458 of the Criminal Code, and IZIMI, the digital vault of the Belgian notariat, to provide for what comes after. We are not building an island: we plug into what the country has already deployed.

5

The capture device

Two speakers facing each other, a room that echoes, a voice that drops when the subject turns difficult: a laptop microphone is not enough. We are working on the design of a capture unit whose specification we write ourselves — it is what will make speaker separation possible without an online service. We are not only a software publisher.

·

The engines, named plainly

Open-source speech recognition and language model, run locally. Desktop application, encrypted database on your machine. Change the engine: the five building blocks remain, and so does the product.

What we build with you

Agility Program

A vendor with thousands of customers sets its priorities without them. We can still set ours with you — and we make that a programme, not a good intention.

1

The beta phase

Four to six trial sessions with the first practitioners, on constructed cases only — never on real patients. Free, no commitment. That is where the corrections come from.

2

The window stays open

The programme does not end with the beta. Any user can propose a change at any time, and we can consider implementing your own requirements in a later release.

3

Who decides

Every proposal is reviewed by a mixed committee: Audioris and practitioners, widened to other disciplines when the subject calls for it.

·

What we promise

A reasoned answer to every proposal, and publication of what was retained. Not a deadline: we promise only what we can hold.

Roadmap

Where we stand, and what follows

This schedule promises only what is funded. What is not in it is not planned.

Today — the product works

Local transcription, structured anamnesis report, identity card reading, database encrypted at rest. Security work is still under way, and we say so without hedging.

Autumn 2026 — pilot phase

Four to six trial sessions with the first practitioners, on constructed cases only. Free, no commitment. This is where the corrections will come from.

2027 — commercial opening

Available to psychologists and neuropsychologists in French-speaking Belgium, on macOS and on Windows.

Then — the other professions

The Police and Corporate verticals exist as prototypes. They are waiting for the Psy vertical to fund itself: we will not open them before.

Beyond

What if Audioris thought with you?

Beyond transcription and reports, we are building the next generation of Audioris: a true intelligent companion woven into your professional everyday — always local, always under your control.

💡

A companion that answers

A technical, clinical, deontological question — or even one about current events? Audioris answers, with citations, strictly within the scope of your discipline. The power of a search engine, calibrated for your profession.

📅

A companion that organizes

Imagine Audioris orchestrating your calendar, preparing your consultations from your case files, managing your appointments — and freeing your mind for what really matters: your patients.

📨

A companion that communicates

Intelligent email triage, drafts written in your style, follow-up on calls and reminders: the administrative load fades away. You remain in the driver's seat — without the mental burden.

This vision is our direction for 2027 and beyond. Every brick of Audioris today is laying the ground for tomorrow's ecosystem.

Our mission