Selector – Like Spotify, But for Your Network

At Networking Field Day 30 we heard from a brand new company, and it was an absolute privilege to be a delegate for this one!

Event Correlation is about more than just Time

As the conversation kicked off they started talking about correlation. Now, I don’t know about you, but whenever I hear the word “correlation” my mind always immediately thinks of time. So, I was immediately thinking that Selector was like a log collector and that it just tied events together based on time stamps. Boy was I wrong!

As Nitin dove deeper into Selector and started to show us its inner workings Selector works just like Spotify or Netflix recommendations – in fact, that’s where they got the idea in the first place!

Those services both take the movies or songs you like and look for the same or similar attributes in other movies and songs to make good recommendations to you. The algorithm might look at genres, artists, actors, instruments, year, and many other things. These things are actually called tags.

It’s all about the Tags (as well as the journey)

Tags might look a little different in the networking world, but they work just the same. Tags might be the IP address and subnet, interfaces, and the user making a change. When a problem is identified Selector looks at all the affected devices and then the tags they have in common. From all this data they can correlate all the tags an event has in common and make some very accurate recommendations on the device to look into for troubleshooting.

Pouring through all the event data, telemetry, and then making the correlations is known as a process called “Observability.”

Why is Observability so Important?

Observability is incredibly important because most systems will simply see an error logged and open a ticket based on a single event. But with the correlation Selector quickly makes the connection that a single event on one device is actually tied to another event that happened on a different device. Now, Selector can submit one ticket for the correlation of several events.

This really turns down the noise from monitoring systems! This is a game changer for monitoring – and that’s whether you’re an enterprise monitoring your systems, or a NOC monitoring all of your customer’s systems!

Final Thoughts

Nitin and Deba really did a fantastic job with this presentation, and this was Selector’s first trip to Networking Field Day! They really knew their audience, they know the problem they are solving is a big one, and they knew exactly how to simply explain how the solution works to a broad audience. Every delegate in the room was just in awe of the technology, the presentation, and the presenters! I am so excited to continue following Selector, and I hope one day I get a chance to work directly with the technology!

To learn more about Selector AI you can:
Watch the NFD30 presentation
Follow Selector on Twitter
Follow Selector on LinkedIn
Checkout Selector’s Website