Researchers annotate data with keywords for describing the physical properties that are observed or modeled. For ensuring findability and interoperability of this metadata, the keywords should be machine-readable and adhere to standardized vocabularies or ontologies. The I-ADOPT framework provides guidelines for expressing such keywords in alignment with the FAIR principles; however, transforming commonly used terms into atomic I-ADOPT components remains a highly manual task requiring both semantic and domain expertise. In response, we propose an LLM-based workflow to generate FAIR-compliant descriptions of variables that align with the I-ADOPT Framework.