AI voice assistant:Let it take notes,gain time for patients

The AI voice assistant is a project for office-based practices that is meant to take structured notes during consultations, prepare patient history forms and take in requests at the reception desk. The results are meant to land directly in the existing practice system, without the practice having to change its software. The assistant is already in use at a practice; Scalableloops supports the project with the AI implementation.

Nikolai Schöbel and Jeremias Burger

Co-founders of Scalableloops GmbH. Nikolai Schöbel leads online marketing and AI strategy, Jeremias Burger the AI architecture. Both build AI systems and train teams on them in their own agency work.

What is the AI voice assistant meant to take over?

The assistant is designed for office-based doctors, medical care centres and GP practices with many patient contacts. According to the project description, it listens, takes structured notes, takes in requests at the reception desk and delivers patient history forms directly into the practice system. It is not meant to interfere with the existing practice software.

We will publish results from operation once they are reliable. Until then, this page describes what the assistant is built for.

Which rules apply to voice AI in a medical practice?

An assistant that listens to consultations processes the most sensitive data there is. Four frameworks therefore need to be clarified before it is used.

The project provides the option of running the assistant on a computer inside the practice, so that consultation data does not leave the premises. Which option fits is for each practice to decide with its data protection officer. You can read more about running AI in-house in our article on local AI for business, and about the disclosure duty in our article on AI disclosure under Article 50.

1. 01

   Health data

   Art. 9 GDPR

   Health data belongs to the special categories of personal data. Processing it is only allowed under the exceptions in Article 9(2), for example for the purposes of healthcare.

2. 02

   Professional secrecy

   Section 203 German Criminal Code

   Doctors may disclose secrets to people who contribute to their professional work, such as IT service providers, as far as this is necessary. They have to make sure these people are bound to confidentiality.

3. 03

   Telling patients

   Art. 50 AI Act

   Since 2 August 2026, people must be informed when they interact directly with an AI system, unless this is obvious. This mainly concerns the reception desk.

4. 04

   Medical device

   EU MDR

   Software can be a medical device if the manufacturer gives it a medical intended purpose. An assistant that only documents and pre-sorts is classified differently from one intended to make diagnoses. This distinction belongs at the start of every project.

What can other practices and businesses learn from it?

Even without this assistant, the project offers principles that apply to any voice AI in a business, whether in a practice, a law firm or customer service.

1. 01

   Keep existing software

   Integration

   An assistant that writes into the existing system is more likely to be used than one that makes the team learn a new program.

2. 02

   Plan for approval

   People decide

   Every draft is read and approved by a responsible person before it goes into the record or leaves the business.

3. 03

   Settle the law first

   Define the purpose

   Confidentiality, data protection, disclosure and the medical device question are settled before the first conversation, not afterwards.