Large-scale instrumental projects, regardless of their scientific field, generate a vast amount of information. Missions such as Gaia have required the production of more than XXX thousand pages of documentation since the project’s inception in 1990. This documentation is essential for ensuring the continuity of knowledge on these long-term projects, where teams evolve as members come and go. Added to this are evolving practices and tools : from a 100% paper-based system just 30 years ago, information has gradually shifted to various collaborative tools such as Jira (task management), Redmine (project management), project management software (scheduling, budgeting, etc.), or systems engineering software (specification management and budgeting, for example).
Under these conditions, how can we find the right information in these oceans of data ?
Artificial intelligence is one answer : we are currently designing a chatbot capable of searching through all databases to answer a question. In addition to the answer, the chatbot will also provide a list of links to the sources where it found the information, allowing users to verify the answer and/or download the supporting documents.
This is estimated to save several days of work per person per year.
