It’s been just over a month since Microsoft Build 2023, which makes it a good time to look at how these announcements could impact you and your data. If you didn’t attend as I did, or read what Microsoft has announced in the Book of News, here is the short list of the announcements that may affect your data (these are all conveniently tagged in the Book of News):
- Introducing the new Microsoft Fabric, built for the era of AI
- New Copilot in Power Pages experiences in preview
- New features and updates for Power Virtual Agents
- Advancements to Copilot in Power Apps
Microsoft Fabric
Yes, OpenAI, ChatGPT, and Copilot are humongous areas of discussion, just as we’ve all witnessed since January with the onslaught of press releases that led up to Microsoft Build. The machines are indeed rising, and this time they are leaving a trail of data with them from prompt engineering, chat threads, and chat command executions.
The good news is that Microsoft did what many might not have expected – they announced Microsoft Fabric. You might assume that Microsoft Fabric is a way for customers to view, interact, and collaborate with their enterprise data that is designed in a way that won’t require a room full of data scientists, engineers, and analysts to make it work, right? Wrong. Copilot is there to help accelerate things, but it still does require many different types of data experts to be in the pilot seats.
Microsoft Fabric harnesses the power of PowerBI and Azure Synapse for raw analytics and visualization, Data Factory for transformation and handling, OneLake for storage, introduces Data Activator for good ol’ trigger-based execution, and Purview to catalog and govern that data. Lots of great stuff here that can handle exabytes of data as if it were on a thumb drive, but it still takes a village to make it successful. And let’s not forget the pricing model of “use more, pay more,” which can range from around $250 per month up to an astounding $275,000 per month. Yikes.
Copilot in Power Pages
When we look at how Copilot has infiltrated the Power Platform, there is a pleasant surprise that came with Copilot in Power Pages. For those of us who remember the old ASP web page days in the original ADX Studio code, we can share virtual high-fives as we see how fluid the Copilot experience makes the mundane tasks we labored over back in the early 2000s (and frankly, up to just a couple years ago) – to get a web portal up and running.
With Copilot in Power Pages at the helm doing the grunt work, more and more customers will be exposing their data to external customers to interact with and generate. Portal chatbots will enhance and add more data into our Dataverse environments at a scale we can’t fathom, and yes, our beloved customers will now be able to make mistakes with the data they submit into our Dynamics 365 CE databases. Ouch. Did I mention that the Copilot can also create tables and fields and, frighteningly enough, remove them too? Are we ready for such an easy way to delete a table or a field in Dataverse?
Copilot for Power Virtual Agents
Speaking of chatbots, the new features and updates for Power Virtual Agents announced the ability to generate dialogue and complete actions. In a chat conversation, the Power Virtual Agent can be configured to leverage connectors of all sorts in the Power Platform (pretty much an all-you-can-eat buffet of enterprise data) and dynamically chain actions from it. These actions can even include Power Automate flows and entity extraction. That introduces significant risk to the data our users input, update, and consume. One domino falling sounds like a lot of work to unravel and correct!
Copilot in Power Apps
If you’ve read this far, you knew this was coming. Copilot will hold your hand to help you make better apps faster. It will also allow Power Apps builders to process any Excel file data into Dataverse no matter what form the data is in (good or bad), which can be helpful for the builders but dangerous for the data governance teams. There’s also no documentation for any “undo” buttons either.
Assuming the app-building process is flawless, it will result in lots more apps for us to consume as end users. According to Gartner research and quoted by Charles Lamanna (Corporate Vice President, Business Applications & Platform at Microsoft) himself, “By 2026, low-code development tools will account for 75% of new application development.” Imagine the scale of things if 75% of app development got 50% easier and faster. It’s safe to say there will probably be double-digit growth in the total number of apps that will be out there creating, changing, and deleting data.
So, in a nutshell, the announcements at Microsoft Build 2023 were wondrous and exciting and, oh, so powerful. Thank you, Microsoft, for the vision and execution of some fantastic tooling. We can expect to see many more apps getting built, including externally facing customer portals with virtual agent chatbots that can inherently execute commands. We’ll have data tools on the backend to manage, analyze, and govern the data.
How to protect your mission critical data in the AI era
The thing we tend to forget about in all of this – which is where Own excels – is the fine details of the data. Getting to a single row or field or affected dataset should be easy, especially when you intend to restore it from a loss or corruption event. It definitely shouldn’t take a village to produce or maintain the ability to “undo” mistakes in the data that are guaranteed to happen. However, those data mistakes are guaranteed to happen sooner due to all these apps and Copilot assistance for the users and builders.
With Own Recover, you can restore datasets at multiple granular levels, including table level, row level, and field level, with relationships intact. To hear more about how Own Recover can prepare you for the new tools and accelerated growth Microsoft has positioned you for, contact us below to see how we can help.
.png){kind=logo}
