Data Activator: Automatic Actions on Data in Motion

[rightturntiger]

It seems like we're constantly bombarded by articles telling us how the volume of data we need to ingest is increasing exponentially and, to further complicate things, the speed at which we need to extract insights from that data is increasing just as quickly. While both of those things are often true, I feel like many of these articles depict this real-time ingestion and these near-real-time insights as happening in a vacuum. Once the data-driven insight is passed on to someone with a responsibility to take action on it, the clickbait authors have lost interest and the story ends.

In real life, however, the story doesn't end there. In fact, in real life, there is often a desire (or a demand) for real-time actions to be taken on that real-time data. Having worked in environments where unexpected downtime or other issues could incur fines by the minute, I can confirm that real-time action is often demanded from groups handling operational data. But, it's impractical to have a team of people watching streams of data and taking preventive or mitigating actions instantly. This is where Data Activator becomes part of the data in motion story.

What's Data Activator? It's part of the family of components present in Real-Time Intelligence in Microsoft Fabric. What does it do? It allows you to set various rules and conditions and, when those are met, take automated actions to address the trends or issues for which you're monitoring. Let's take a closer look at what we can do with Data Activator.

One of the things I feel like is most important in a tool that allows us to take automated actions on things is that we can establish a rich set of rules and conditions. Whether we're dealing with monitoring data from Database Watcher or data streaming from IoT devices, we want the ability to create more nuanced rules than "do x if y is over 200". Having supported some transportation and retail clients in the past, oftentimes an initial indication of a problem with equipment can be something like "the temperature has stayed in a certain range for 30 minutes". This is one of the areas, in my opinion, where Data Activator really shines, because it allows to create these rules and conditions in a low-code/no-code, almost conversational way. If you can write a sentence describing the condition for which you're monitoring, it's quite likely that you can create that exact rule in Data Activator.

But what happens when the rules are created? What actions can we actually take? Let's quickly walk through the three types of actions we can take. First, we can send an email if the condition is triggered. Second, we can send a Teams message if the condition is triggered. These are both somewhat passive monitoring actions but are often appropriate for many scenarios. 

The third, and most exciting, action we can take is to trigger a Fabric item. This can be a notebook or a pipeline. While this is where the low-code/no-code experience of Real-Time Intelligence ends, this ability also opens up a vast set of possibilities for actions we can take to handle the situation at hand. 

I hope you've enjoyed this introduction to Data Activator now that it has gone GA at Ignite! I look forward to continuing this blog series with some deeper dives into individual Data Activator scenarios, but for now I encourage you to spin up a Fabric trial and play with Data Activator, and all of Real-Time Intelligence, yourself and discover scenarios where it can help you.