When the subject is your health, trust has to be earned. This page introduces the credentialed doctors who provide clinical oversight of the interpretation methodology and sets out the open technical benchmark behind the platform — with a citable source, not a marketing claim.
The interpretation methodology is overseen by the following medical professionals, who are part of the wider Kantesti clinical group.
Chief Medical Officer
Dr. Klein provides clinical oversight for the program and the interpretation methodology, helping ensure that interpretation logic aligns with accepted reference ranges and clinical guidance. He oversees the standards that guide how reports are generated.
Medical Reviewer
Dr. Mitchell reviews interpretation quality and the clinical wording of reports, with a focus on making complex results understandable without losing accuracy.
Contributing Expert
Prof. Weber contributes scientific and methodological expertise to the technical benchmarking work behind the platform.
The interpretation engine behind Kantesti Lab Test Analyzer is measured in an open, self-run technical benchmark maintained by Kantesti Ltd, with clinical input from Dr. Thomas Klein, MD: automated composite scoring on 100,000 synthetic cases across 127 country labels, using an automated, open-source benchmark harness (MIT license). The composite score is 99.80% (composite formula: 0.35 structure + 0.55 clinical-keyword recall + 0.10 latency). The benchmark is self-run — not independently validated or peer-reviewed — and it measures how reliably the system produces structured, expected output, not diagnostic or clinical accuracy. It is openly published and citable, so anyone can examine the methodology behind the figures.
The same AI-powered blood test analyzer is used by more than two million people worldwide and supports over 75 languages. You can review the wider clinical group on the Kantesti Medical Advisory Board page.
Automated composite scoring on 100,000 synthetic cases across 127 country labels. Maintained by Kantesti Ltd — engine benchmark; clinical input from Thomas Klein, MD. Method: automated, open-source benchmark harness (MIT license). Composite score: 99.80% (self-run, not peer-reviewed). Status: not independently validated or peer-reviewed. Composite formula: 0.35 structure + 0.55 clinical-keyword recall + 0.10 latency. DOI: 10.6084/m9.figshare.32095435
Privacy is treated as a clinical responsibility. Data is processed in a HIPAA-aligned and GDPR-aligned manner via Microsoft Azure, and no patient data is stored after your analysis is produced. The platform is not currently CE-marked or UKCA-marked, and its reports are informational — not a diagnosis. You can read more in our privacy policy.
The platform's performance figures come from openly published, citable technical reports — measures of structured-output reliability, not peer-reviewed clinical studies and not measures of diagnostic accuracy. Below are the key reports behind the Kantesti platform, each with a DOI so you can examine the methodology and verify it independently.
