In the rapidly evolving world of AI, keeping up with the latest tools and technologies can feel like a full-time job. One of the most exciting recent releases from Microsoft is the introduction of AI Skills in Microsoft Fabric. With the potential to streamline operations, boost productivity, and even change the way we work, this technology is certainly worth a closer look.
In this article, I’ll break down what Microsoft Fabric AI Skills does, how it stacks up against Microsoft Copilot, and how you can get started using it to enhance your data operations. By the end, you’ll have a clear understanding of whether this tool is worth exploring further and how it could fit into your workflow.
But before we dive into the details, here’s a video I recently recorded that walks through the process step by step:
What is Microsoft Fabric AI Skills?
Microsoft Fabric AI Skills is part of the growing portfolio of generative AI technologies that Microsoft has been adding to its ecosystem. In essence, it allows you to run natural language queries directly on top of your data assets, such as a lakehouse or warehouse. The big advantage here is that AI Skills generates T-SQL queries, making it a powerful tool for analysts who need to dig deep into datasets.
However, unlike some other tools like Microsoft Copilot, which generates responses based on DAX, Fabric AI Skills focuses on generating SQL queries, providing more direct insights from your data.
Getting Started with AI Skills
One of the best parts about AI Skills is how simple it is to get started. If you’re an administrator, all you need to do is enable it in your tenant settings, and you’re off to the races. From there, it’s just a matter of navigating to your workspace, selecting a lakehouse or warehouse, and clicking on the new item to start working with AI Skills.
The interface is straightforward, allowing you to ask natural language questions like, “How many customers do I have?” or “What were my sales in 2022?” and receive a SQL query that provides an accurate response.
In the video, I demonstrate this by running a query on customer data and another on product sales. The AI Skills tool not only picks up on nuanced questions but can also be trained with custom prompts to generate even more refined results over time.
Copilot vs. AI Skills: Who Wins?
Throughout the video, I also compare Fabric AI Skills with Microsoft Copilot. While both are incredibly useful tools, Copilot still has its limitations—especially when it comes to complex data queries. Copilot can sometimes generate nonsensical charts or miss nuances in queries, whereas AI Skills tends to get it right, especially with the ability to train it further with example queries.
AI Skills also allows for the inclusion of model-level notes to guide SQL generation, making it much more adaptable for specific business contexts.
Tragic Flaws and Areas for Improvement
No technology is perfect, and Microsoft Fabric AI Skills is no exception. In the video, I discuss some of its shortcomings that are worth mentioning:
- High Cost of Entry: AI Skills requires F64 capacity or higher, which can be costly—around $9,000 a month.
- SQL-Only Support: Right now, AI Skills only works with SQL endpoints like lakehouse or warehouse. It doesn’t work with semantic models, which can feel like a step backward for organizations that have already invested heavily in building semantic layers.
- Lack of Conversational History: Unlike ChatGPT, AI Skills doesn’t yet retain conversational history. This can make refining your queries difficult since the tool doesn’t remember past interactions or ask for clarifications.
- No Instant Feedback Loop: When AI Skills generates suboptimal SQL, there’s no built-in way to fix the query immediately. Instead, you have to manually tweak the query, run it elsewhere, and then come back to update your prompts—something that could be streamlined.
Final Thoughts: Is This a Game Changer?
While AI Skills is not without its flaws, it’s a promising tool with great potential. For those who are willing to invest the time in training it with custom queries, it can become an invaluable asset in your data toolkit. However, the high cost of entry and the lack of conversational history are significant drawbacks that Microsoft will hopefully address in future updates.
In comparison to Microsoft Copilot, AI Skills shines in its ability to generate more accurate SQL queries, though Copilot has its own strengths when it comes to visual data manipulation in Power BI. If you’re already working within the Microsoft Fabric ecosystem, I highly recommend exploring AI Skills to see how it can enhance your data-driven decision-making.
So, is it time to change your career? Probably not. While AI Skills is impressive, we’re not quite at the point where AI will take over all aspects of data analysis. Instead, it will continue to help us do more with less and become more efficient in our roles.
Let’s keep an eye on this space and see how things evolve in the coming months!
business intelligist 