Did you attend FabCon during the last week of March in Las Vegas, NV? If you were not able to attend, it was packed with a lot of great knowledge, exciting announcements, and wonderful data professionals to bounce ideas off of all week. I was lucky enough to attend with my colleague and friend, Michael Heath. At every break when we would reunite, we were so excited to share what we had just learned.
I know not everyone got to attend this event. And even for those of us who were there, it was a lot of information to take in over three days. I wanted to summarize some of my top takeaways. I mean, it is difficult to distill an entire conference into a single blog post, but I wanted to at least highlight some of the key moments that made me go aha!
FabCon Takeaway #1 – Out of the Box
While I heard this many times before, I finally was able to wrap my head around it during the conference. Fabric is an all inclusive platform to simplify the management of data for all forms of projects. Whether it is warehousing, reporting, or data science, the platform is ready to use.
Like Buying a Car
The best analogy I heard all week was the concept of buying a car. Before Fabric, you needed to pick out each part individually. You needed to identify your storage medium, ETL tool, virtual network components, and data science tools. This is the equivalent of going to the auto parts store and picking each component out one by one and then assembling it to build a car. While some of you might choose to do this, it is a lot of work. You need to not only have the skills to do the data work, but the experience to setup your Azure environment.
Instead, Fabric allows you to buy the car as it is off the lot. When you setup your workspace, you can add components without the hassle of configuration. It just works which is a really nice feeling. Even better, because you are using a capacity license, it is easier to estimate costs. I know there are calculators out there for Azure, but you always forget something when you are attempting to build a budget. Fabric eliminates that challenge with a clear pricing structure. This is a huge step forward for self-service data science outside of the traditional IT function.
Microsoft has been pushing their “5×5” approach with Fabric. The concept is 5 seconds to sign up and 5 minutes to your first wow. While I appreciate this approach, I think there is a little work to go in this space. If you have a guide to help you, I believe this to be true. However, if you are new to Fabric, I do not find it to be quite intuitive for that. Long story short, make sure you do a little learning before you try it on your own.
What About Data Engineers?
Does this mean they are out of a job? Not at all! Data engineers will still be essential. For me, I think it redirects their efforts into more valuable workloads. Instead of fighting with Azure, they are focused on getting value out of the platform. It also opens the door to data engineers and scientists with deep domain knowledge in the business. That is a game changer!
Does this mean our existing data infrastructure is obsolete? Also no! These tools are still essential. If anything, this might allow less important workloads be shifted to Fabric reducing administrative burden. That allows you to focus on the mission critical workloads within your existing infrastructure. While Microsoft would like you to move your workloads to Fabric, not everything has to go.
FabCon Takeaway #2 – CI/CD Is Close
If unfamiliar, CI/CD stands for Continuous Integration and Continuous Deployment. To handle this, you need to have a mechanism to manage your source code for components of Fabric. While it is not 100% there yet, the future is quite bright. While some might not like how this is done, this is a huge step forward.
The integration with Azure DevOps is slick. The ability to create a feature branch and merge it back to the main branch is slick. With a few clicks, you can create a new workspace, replicate your components from the main branch, and replicate it to Git. When you are finished with your feature, you can easily merge it back in. But for me, the nicest part is that you can retire your workspace as part of the merge which is so nice.
I also love the fact you can choose your deployment path. If you would prefer to use Deployment Pipelines in Fabric, you can do so. However, if you rather have Git repos for Dev, Test, and Prod, you can do it too. This allows you the flexibility to manage and control your deployments as you see fit.
Stay tuned for an article on this in the near future!
FabCon Takeaway #3 – Integration Points
For me, the biggest question I kept asking was how well could I integrate these components into other solutions. Creating a machine learning model is fantastic, but useless if you cannot integrate it with other applications. This is huge!
I need to investigate how this works further, but the promise was pretty clear. The idea is that your data scientists can perform their analysis and train a model with Microsoft Fabric. Once completed, they can leverage it via endpoints and provide predictions. I have done this with Azure ML in the past, but the process was a tad clunky. I am hoping this experience gets better with Fabric.
Another teaser was the ability to build your own custom Copilot chat bots within Fabric. This is not using Copilot to create a model, but rather training a Copilot to use your data. Eventually, this data could find its way into other tools. To be honest, this has been somewhat happening already today with the Q&A feature in Power BI. However, if Microsoft can streamline this into a single Copilot experience, this would be huge. Being able to query your data without having to go Fabric would go a long way.
Anything Else from FabCon?
Yes, there is a ton more to talk about. But these were the stand outs for myself. What is hard to put into a post are the conversations with other professionals were outstanding. Meeting some of the conquering heroes of the Power BI world is always fun, but eating lunch with different people was where the biggest insights were gained. Learning about everyone’s unique scenarios allows you to refocus your vision of the platform. I always recommend people attend conferences for that reason alone. That is where you gain the most interesting insights!
Also, since everyone asks this question, Michael and I were able to enjoy some of the fun around us. It was a busy week at the conference and work did not slow down in our absence. While we did not get to any shows, we did get to unwind with other professionals and even meet one of the best Elvis impersonators I have seen in my lifetime.
Conclusion
This coming year is going to be exciting when it comes to Microsoft Fabric. Between what was announced at FabCon and what is coming in the near future, the future is bright! I look forward to sharing more insights over the coming months.
Did you make it to FabCon? What were your key insights? Anything I missed in my summary? Tell me in the comments below!