Averbis Medical Summary

A Freiburg language-processing company since 2007

Averbis GmbH is based in Freiburg im Breisgau and registered at AG Freiburg under HRB 701080 (averbis.com/de/impressum). Philipp Daumke and Kornél Marko founded it in 2007 and are its managing partners, with Patrick Oestringer as managing director (averbis.com/de/unternehmen/leadership). In 2018 the business moved from services to products (averbis.com/de/unternehmen/ueber-uns). The company calls itself a specialist in language processing and large language models in healthcare for more than 15 years.

The homepage claims more than 100 customers and partners and shows logos including UK Freiburg, IQVIA, CGM and Med Uni Graz. That count is the vendor's.

What does Medical Summary produce?

Summaries are "based only on existing documents", each with source references and direct links to the original text (averbis.com/de/produkte/medical-summary). A dialogue function lets staff query the documentation in plain language.

Discharge-letter drafts are assembled from findings, operation reports, diagnoses, and nursing and therapy documentation. A physician reviews, completes and approves the draft. The named inputs are the universal archive, the KIS, documentation systems and the national records ePA, ELGA and EPD.

Medical Dialog adds transcription of doctor-patient conversations and structures diagnoses, medication and vital signs.

Hosted in the hospital, with optional EU cloud components

The product runs inside the hospital IT. Optional cloud components process data temporarily, only on servers in the EU or Germany, at providers with ISO 27001 certification and a C5 attestation. Those certificates belong to the providers, not to Averbis. The company claims GDPR conformity for the external AI components and names no language model.

Integration with the KIS goes through "standardised APIs". Two homepage testimonials, from Meierhofer AG and ÜBAG Hausarzt-Zentrum, mention deep KIS integration and transfer into a practice system. The vendor chose both. The product page does not mention MDR or a CE class.

Four papers on the text-mining base, none on Medical Summary

PubMed lists four papers with Averbis in the title or abstract, and all concern the company's text-mining technology. They include a 2024 de-identification pilot for the German GeMTeX corpus (PMID 39234720) and a 2022 paper on mapping language-processing annotations to FHIR (PMID 35672965). The About page claims involvement in more than 100 publications. No study of Medical Summary itself is listed on the vendor pages.

Price

Averbis posts no price. Hospitals request a demo through the contact form (averbis.com/de/produkte/medical-summary).

Who is it for?

The inputs it works from, findings, operation reports, diagnoses and nursing and therapy documentation, are hospital records, and the page names hospitals as the ideal customer. HIS vendors are the other buyer in the company's customer logos. The product pages and the named record systems are German-language.

The verdict

Medical Summary summarizes and drafts letters from existing hospital documents and runs on the hospital's own servers. Visit transcription is a separate module, Medical Dialog. No study of Medical Summary itself is listed on the vendor pages.