On-Premise AI Avatars
for US Hospitals

A full-body AI avatar or hologram for routine patient conversations, on your servers, with no internet or cloud. Interactive, real-time, full-body, on-premise or fully air-gapped.

No Patient Data Leaving Your Hospital.

Routine conversations eat into time with patients

Questions and paperwork fill the hours your teams should spend with patients.

Staff shortages

Teams are stretched thin, and the same patient conversations still take their time.

Waiting times

Patients wait weeks for a first consultation, while routine questions fill every visit.

Admin burden

Filling out medical documents adds to the daily workload of clinicians and staff.

What is an on-premise AI avatar for hospitals?

A real-time, full-body AI avatar that handles the patient conversations that don’t need a clinician, and keeps patient data under your control.

  • Real-time voice and video. A face, a voice, and natural turn taking. Voice only where video isn’t wanted. Runs in the browser, no app to install.

  • Any screen a patient meets. Kiosk in the waiting area, Holobox at the entrance, and web, mobile, or telephone with the on-premise option. One persona on every screen.

  • Private by design. Runs on your own servers, with access, AI models, and updates under your IT team’s control.

  • Answers only from your content. The avatar reads from forms and material your clinical team approved. Nothing is trained on patient data.

It doesn’t diagnose, interpret symptoms, triage, or prescribe.

Start with intake. Add roles when you’re ready.

The first role is patient intake. The same avatar can take on more roles later, one at a time.

AI patient intake before the visit

AI patient intake replaces the paper form with a conversation. An avatar on a kiosk, screen, or hologram talks with the patient in their own language, by voice or text, and collects what a hospital needs before care starts: who the patient is, why they are coming, and what the care team needs to know about their health.

It asks every patient the same questions in the same order, and follows up when an answer is unclear.

Talk to Sales

The avatars provide information, not medical advice. When a patient needs a clinician, the avatar hands the conversation over to your staff.

Other roles on the same platform: Front desk, Navigation, Education, Follow-up, Staff training.

One visit, from intake to final notes

Each step prepares the next, so the doctor starts informed and the notes are drafted.
Clinicians get more time to devote to patients and less time on paperwork.

  1. 01

    Patient completes intake

    The avatar takes the patient through your intake form, in their language.

  2. 02

    Doctor gets a summary

    The summary reaches your system, and the doctor reads it before the visit.

  3. 03

    Visit notes get drafted

    A device in the room drafts the notes, and the doctor approves them.

What changes for your staff and patients

For your hospital

Lower staff workload

The avatar takes the repetitive conversations, so staff time goes to the patients who need a person.

More capacity, same staff

One avatar talks with many patients at once, and each kiosk or hologram adds more.

Consistent answers

Patients get answers from content your clinical team approved, the same on every channel.

For patients

A face-to-face conversation

Patients talk to a lifelike avatar that listens and answers. A more natural way to get help.

Attentive service every time

The avatar stays calm and patient in every conversation, however busy the front desk gets.

Help without waiting

Patients get information and complete routine steps right away, without queuing for the front desk.

Where Cliav sits in your hospital

Avatar, speech, and language models run on your own hardware, under your control.

Your hospital

Your Users

Patients

Intake, check-in, scheduling, and follow-up in their language

Staff & Clinicians

Intake summaries before the visit, and training practice

Cliav AI Avatar on a screen, kiosk, or hologram

Behind the Avatar

Speech models

Listen and reply in their language

Language models

Open-weight, with no outside API

Approved content

Your forms and clinical material

Hospital systems

Receive the intake summary

Your IT team controls access, AI models, and updates.

Air-gapped

No internet connection. Patients use Cliav on site, and it connects only to internal systems you choose.

On-premise

Your servers, connected as your IT team allows, for online intake, follow-up, and reminders.

Open integration interfaces

REST API, webhooks, WebRTC, and WebSockets.
EHR connection is scoped in your pilot.

LLM agnostic

Runs on the language model your hospital approves, including open models on premise, with no vendor lock-in.

Who’s behind Cliav

Cliav is the US line of RAVATAR, whose avatar platform has paying customers in four sectors, including PwC, Capgemini, and AMD.

Built by RAVATAR

The avatar platform behind PwC, Capgemini, and AMD projects, with permanent installations in Florida.

Paying in healthcare

A public healthcare IT provider in Italy has run our avatars on subscription since January 2026.

Letters of intent

11 letters of intent from hospitals, universities, and patient organizations in Europe.

Founders

Andrii Bidochko, CEO

Built the agent evaluation engine that a national defense agency runs in a closed network.

Artem Synytskyi, CTO

Built the on-premise avatar pipeline with Intel, and a clinical notes system for clinics.

Oleg Avershyn, CMO

Won PwC, Capgemini, and AMD, and signed AVI-SPL as RAVATAR’s US reseller.

Questions from hospital review teams

Does Cliav need an internet connection?

Not in the air-gapped option: only the internal systems you choose. On-premise connects as your IT team allows.

Where do the models run?

On site. Open-weight models run on your own servers, so your hospital owns its intelligence.

Who controls access and updates?

Your IT team. Cliav runs on your own servers, so access, AI models, and updates stay under its control.

Who approves the clinical notes?

The doctor. A device in the room drafts the notes, and the doctor approves them.

Do patients know they’re talking to an AI?

Yes. Cliav tells patients it’s an AI, and an on-screen notice says so too.

How does a pilot work?

Pick one role and one location, load your approved content, run it for an agreed period, and review the analytics.

Does Cliav replace clinicians?

No. It handles routine conversations. It doesn’t diagnose, interpret symptoms, triage, or prescribe.

Ready to keep your AI inside your walls?

Cliav runs on your own servers, with no outside AI API.