AI-assisted pharmacovigilance

Our team built almost the entire platform, maintained and extended over four years:

• Managing clinical studies: creation, duplication, associated files, sponsors and categories.
• Integrating the MedDRA dictionary: importing successive versions and a search engine that navigates the hierarchy, down to the lowest level terms.
• Coding verbatims: bulk import, editing, automatic suggestions, exact matches and search conventions.
• Coding medicines through WHODrug: active substances, B3 and C3 suggestions, countries and a dedicated search.
• Upgrading dictionaries from one version to the next, with an impact analysis showing what the change means for existing data.
• A query circuit between data managers: built-in messaging, opening and closing requests.
• The exports: verbatims, WHODrug, universal export, coding trail and client exports.
• AI-assisted coding, powered by our own engine, which automatically proposes the most likely medical terms for each verbatim, together with real-time notifications between users.

A demanding, regulated field, healthcare, in which two things had to be delivered at once: a robust business platform, and the artificial intelligence engine that sets it apart from its competitors.