Controlled vocabularies ensure consistency in wording and spelling. They alleviate ambiguities and misunderstandings while placing limits on content entry. Without them, it is hard to develop tools, perform data integration, etc. It is not without a reason that they are deeply embedded in FAIR data principles and the Semantic Web. But how to build them and serve them? Definitely not as static PDFs, which sometimes you have to buy (e.g., from large standardization bodies such as IEC). In this talk, we will present an effective and open-source pipeline for developing and serving controlled vocabularies.