A private AI workspace for your medical practice

Picture a doctor pasting a patient's history into a free chatbot to speed up a referral letter. Useful, until you ask where that history just went: out of the practice and into a cloud you have no agreement with, attached to a real person's health.

Your practice can use AI for the routine work, letters, summaries, drafting, without handing patient data to a third party. It depends on where the model runs and who controls the platform around it.

Patient data is the most sensitive data there is

Health information is exactly the kind of data that is not meant to wander. A public chatbot moves it onto someone else's servers the moment a note is pasted in. You feel that sharper than most businesses, because the trust between you and a patient rests on it.

Telling staff "do not use AI" rarely sticks. They reach for the quickest tool unless you give them a safe one.

Keep the model inside the practice

With kral, everything runs on your own server. Add a local model on your own hardware and a note about a real patient goes to your machine and no further. Nothing leaves the practice, because there is no external service in the path. Many practices mix it: a cloud model for general text, a local model for anything that names a patient.

A real workspace, not just a chat

Your team can build their own assistants without any code: one that drafts discharge letters in your house style, one that turns rough notes into a clean summary. Save them as routines, so the next person does not start from scratch. Drop in a document and ask about it, get a current cited answer from the web when you need one, and switch between the leading models in a click. It all lives in one place behind your login.

Connect your own records and templates

kral supports MCP, the open standard for connecting tools and data to an AI. The assistant can draw on your own letter templates and internal knowledge through a connector you control, rather than guessing from the open web. Your records stay in your systems.

You hold the keys, and you see the usage

Decide who uses which models and cap what each person can spend, so there are no surprises on the invoice. Watch real usage on a dashboard, and let staff sign in once through single sign-on. It installs on Windows Server behind IIS, the same idea as a company-wide AI you host yourself, stays inside your network behind your firewall, and wears your practice's branding.

And we help you set it up

You do not have to do the technical part alone. We install kral with you, connect it to your tools and advise on bringing AI into the practice without patient data leaving your side. Implementation is part of what we offer.

Give your team a real AI workspace and keep patient confidentiality where it belongs. See it running, then let us help you put it on your own server.

Book a demo

Open the app