Actors in the field of ferments and fermented foods need more comprehensive databases on their topics (e.g. metabolic networks of organisms, microbial interactions, ...). Much of the information of interest is already disseminated in scientific publications, and each need could be addressed by a specialized information extraction solution, which can extract and structure it in a database. But there are as many specific needs as there are different actors, and even for the same actor, their needs may also evolve over time. We propose the development of a web platform allowing the collection and formalization of actors' needs through a user-friendly interface. The platform will execute a robust information extraction pipeline in order to automatically create a specialized database and a query/visualization interface from a corpus of texts. Additionally, the platform will take into account possible user corrections. Therefore, the project’s objective is to reach a prototype (TRL4-5) by creating databases of interest for the different partners involved in this project.