AI in SQL Server 2025

Most teams that want to search their data by meaning end up bolting a separate vector database onto their stack, then paying for the data movement, syncing, and governance headaches that come with it.
Most teams that want to search their data by meaning end up bolting a separate vector database onto their stack, then paying for the data movement, syncing, and governance headaches that come with it.
In this course, AI in SQL Server 2025, you'll gain the ability to build semantic search directly inside the database you already run, using nothing but T-SQL.
First, you'll explore what SQL Server 2025 adds for AI workloads, including vectors, embeddings, and the new functions behind them, and learn when these features are the right fit.
Next, you'll discover how to generate embeddings from your text with an external model, and store, compare, and validate them in a vector column.
Finally, you'll learn how to assemble an end-to-end semantic search workflow by chunking long text, querying by similarity, and scaling with vector indexing, then see how to operate it responsibly with the right tooling, constraints, and monitoring in mind.
When you're finished with this course, you'll have the skills and knowledge of AI features in SQL Server 2025 needed to add meaning-based search to your own catalogs and knowledge bases without leaving SQL Server.
Homepage

---