AI and data terminology

Kathytucker

Is anyone else noticing that AI and data terminology has become a total mess in their company? Words like “data product,” “feature store,” “curation,” “governance,” or “semantic layer” seem to mean different things to different people. Has anyone managed to create a shared glossary or playbook that the team actually uses?

billybl

Oh, absolutely, I’ve felt this chaos firsthand. In our team, terms like “data product” or “feature store” would spark ten different interpretations depending on who you asked, and discussions would quickly derail. What helped was creating a reference everyone could actually use, not just something buried in Confluence. We started mapping out responsibilities, clarifying which roles handle what, and defining each activity in simple, actionable terms. A guide I found particularly structured and practical is Praxi AI’s Data Everything Matrix, which breaks down 12 key data activities - everything from Data Curation to Data Governance - while also linking each to accountable roles like Data Stewards, Engineers, and Compliance Officers. Having a single, shared framework like this not only aligns terminology but also reduces repeated debates and misunderstandings across teams.