AI-Assisted Development in Microsoft Fabric: Hands-on Training for BI and Data Teams

AI coding agents can speed up Microsoft Fabric development considerably. Used carelessly, they also produce plausible-looking pipelines, models and DAX that fail quietly. This 2-day, hands-on course shows you how to get the benefit without the risk. You will learn a repeatable way of working: plan first, build in small controlled steps, verify everything, and capture your team’s standards so the agent follows them every time.

Across two days you will connect an agent to a sandbox Fabric workspace, plan and build a medallion lakehouse and semantic model, and then deliberately hunt for defects in AI-generated work. You leave with a working method, a first house skill, and a standards roadmap for your own team.

Develop AI-Assisted Development Skills:

• How agents work (context, tools, skills and MCP), and where AI helps and hurts in BI development.

• How to set up a safe environment, with a dev workspace, Git integration, least-privilege access and a read-only first connection.

• How to manage context effectively, avoiding context rot and breaking work into small tasks with fresh sessions.

• How to write a buildable plan covering design, tasks, acceptance criteria and out-of-scope items, and have it reviewed and approved in DevOps.

• How to build lakehouses, notebooks, pipelines, warehouses and semantic models from a plan, with items in Git and checkpoint commits.

• How to use a modelling MCP and review generated DAX, and how to prepare models for Copilot and data agents.

• How to verify AI output using tests, reconciliation and review checklists for SQL, PySpark and DAX, including checks for secrets, PII, lineage and capacity impact.

• How to handle change by updating the plan rather than patching the output, spotting drift, and promoting through deployment pipelines.

• How to write, test, version and own “skills” that codify your team’s naming, layering and modelling standards.

Learn AI-Assisted Development Your Way, Training Tailored to Your Needs

This course is available in a UK-based face-to-face classroom or virtual setting and can be tailored to your needs. If your team has particular tooling, governance requirements or Fabric architecture in place, let us know and we can focus the exercises on your environment.

Solve Your AI Development Struggles with Expert Training

Whether you are a BI developer, data engineer or analytics lead, you are probably asking:

• How do I use AI agents to build faster without lowering quality?

• Why does the agent get worse the longer a session runs, and what should I do about it?

• How do I stop an agent touching things it shouldn’t?

• How do I review AI-generated SQL, PySpark and DAX quickly and reliably?

• How do I get the whole team’s AI output to follow the same standards?

If this sounds familiar, this is the course for you. We will walk you through practical answers to each of these questions, with exercises you can take straight back to your own projects.

This course is aimed at people who already build in Microsoft Fabric. Delegates should have:

• Working knowledge of Microsoft Fabric, including lakehouses, notebooks, pipelines and semantic models, or equivalent experience in Power BI and Azure data platforms.

• Familiarity with SQL, and a basic understanding of PySpark and DAX (you will be reviewing these, not writing them from scratch).

• A basic understanding of Git (commits, branches and repositories) and of Azure DevOps or a similar work-tracking tool.

• Some familiarity with the medallion architecture (bronze, silver and gold layers) is helpful but not essential.

No prior experience with AI coding agents is needed. Each delegate will need access to a Microsoft Fabric sandbox workspace, which we can provide or help you set up. If you are unsure whether your experience is the right fit, get in touch and we will be happy to advise.