Creation of a virtual assistant guiding plant researchers in making their datasets FAIR-compliant

Plant phenotyping datasets are highly heterogeneous and difficult to annotate consistently. This project will develop a conversational AI 'virtual assistant' that extracts structured metadata from natural-language descriptions and produces ISA-JSON and RO-Crate outputs. The assistant integrates with ISA tools and reduces the burden of metadata creation, improving FAIR compliance across plant research communities.

Predicted outcomes:

• Conversational AI service for extracting structured metadata
• MetaBuddy web app and integration into ISA Wizard
• Standards-compliant metadata outputs (ISA-JSON, RO-Crate)
• Improved metadata quality for plant phenotyping datasets
• Sustainable, open-source software for long-term community use