AI for SQL Server

Copilot in SSMS, Azure OpenAI in T-SQL and vector search in SQL Server 2022: the AI options around your database have matured in a short space of time. In this course you apply these techniques to your own data and build a complete AI layer on top of an existing SQL Server.

Description

SQL Server has gained a substantial set of AI capabilities. Copilot in SSMS (preview) generates queries based on natural language, explains existing T-SQL and suggests indexes and query rewrites. In this course you learn to put those features to genuinely productive use, not as a toy but as a day-to-day working environment.

You also connect Azure OpenAI to your database. From T-SQL you call an endpoint, use Machine Learning Services for R and Python and work with the new vector data types in SQL Server 2022. Storing embeddings, running similarity searches and combining results with classic SQL: it is all covered.

You practise in a Learnit test environment with SQL Server 2022 and a connected Azure OpenAI endpoint. That way you work with real data and real AI calls, without first having to set up your own infrastructure.

To finish, you build a complete use case: classifying customer tickets, determining sentiment on text fields and making documents searchable with vector search. Security has a fixed place in this, from preventing prompt injection to shielding data outflow.

Topics

• Copilot in SQL Server Management Studio (preview): generating queries in natural language
• Having existing T-SQL explained and complex queries described
• Optimisation suggestions: index recommendations and query rewriting via Copilot
• Asking schema questions: having tables, columns and relationships found for you
• Deciphering error messages with help from Copilot
• Connecting Azure OpenAI to SQL Server: configuring the endpoint and calling it from T-SQL
• SQL Server Machine Learning Services: R and Python within SQL
• Vector search in SQL Server 2022: storing and searching embeddings
• Practical applications: document search, customer segmentation and sentiment analysis on text fields
• Building a complete use case: combining SQL data with AI classification for customer tickets
• Security: preventing prompt injection and shielding against data leakage
• Setting up audit trails for AI calls from SQL Server

Result

After this course you will work productively with the AI capabilities in and around SQL Server. You use Copilot in SSMS as a fully fledged working partner, connect Azure OpenAI to your database and add vector search to existing tables.

Specifically, by the end you will be able to:

• Using Copilot in SSMS to generate queries, have T-SQL explained and request optimisation suggestions
• Resolving error messages and schema questions efficiently with help from Copilot
• Configuring an Azure OpenAI endpoint and calling it from T-SQL
• Using Machine Learning Services to run Python or R within SQL Server
• Storing embeddings in SQL Server 2022 and running vector search on text and documents
• Classifying customer tickets, determining sentiment and making documents searchable
• Recognising and preventing prompt injection, shielding against data leakage
• Setting up audit trails for AI calls so that you remain demonstrably in control

Target audience

The course is aimed at experienced SQL Server professionals who want to add the new AI capabilities to their toolkit. You work with SSMS and T-SQL on a daily basis and want to know how to apply Copilot, Azure OpenAI and vector search to your own databases.

The course is intended for:

• SQL Server database administrators (DBAs) who want to increase their productivity with Copilot
• Data engineers who want to add AI functions to their ETL and data platform
• BI developers who want to carry out sentiment analysis or classification on text fields
• Application developers who use SQL Server as a backend and want to integrate AI features
• Data architects who want to work strategically with AI functions in SQL Server

Teaching method

The course is entirely hands-on and technical in design. You work in a Learnit test environment with SQL Server 2022 and a connected Azure OpenAI endpoint, so you can get started straight away with real data and real AI functions. Short explanations alternate with practical assignments in which you build your own queries, embeddings and classifications.

Because of the technical depth, we work in small groups of no more than 6 participants. This gives you room for your own questions and personal guidance from the trainer, including on points that are specific to your database environment.

You receive a workbook with a dataset and step-by-step plan, a reference document with code examples and a certificate of participation. By the end you will have delivered a working AI use case, including vector search and classification on customer tickets.

Prior knowledge

You have solid T-SQL skills, gained in our SQL Advanced course or through equivalent practical experience. You are comfortable working with SQL Server Management Studio and know the basics of relational databases. Prior knowledge of Azure or OpenAI is not required; we cover those from the ground up. Experience with Python is an advantage, not a requirement.

Comments

You work on a Learnit laptop with SQL Server 2022 and a connected Azure OpenAI endpoint. Bring your own laptop if you prefer, but this is not necessary.

In-company

Would you like to organise this training for a whole group or several groups? This can be cost-effective. For example, we can tailor the content entirely to your organisation, and training a whole group is also cheaper.

Are you looking at large numbers of participants or a complete academy? That involves a great deal. Consider planning and communication, for example, or identifying and agreeing the objectives and content of the training. Or differences in level, feedback mechanisms and safeguarding. Fortunately, Learnit has extensive experience in this field (more than 25 years) and we are happy to help you through this process.

Request a quote

Language

The training is given in Dutch as standard. The trainer has a good command of English. English-language course material can be used. With a minimum enrolment of 3 participants, the training can also be given entirely in English.