Aug 21, 2026 · 46 min · 13 segments
Network infrastructure generates enormous volumes of data, but much of it remains difficult to access, interpret, and use. In this episode of Data Unchained, John Capobianco, Head of AI and Developer…
John CapobiancoGuest
Molly PresleyHost
And kind of tying back to the data portion of the conversation, since we're on data unchained, you and I were chatting a little bit when we first met about how networks generate massive amounts of data, but that remains really difficult to use, difficult to access.

Is that kind of an area that Itential is digging into, or is it more of an industry-wide problem?

I think it has been and really it exploded with the things, you know, the more capabilities the network started to carry when we get into voice and we get in over power, over Ethernet and we get over wireless and all of the hyper connected things and then maybe Internet of Things.

telemetry is trying to solve this problem, right? So you can build agents that can stream in that telemetry and help you make sense of it in human language, right? So I think that some of the barriers in the networking industry in particular around accessing the data are going to collapse.

I think that if we start to shift between humans interfacing with Originally, directly with the devices, and then more and more towards controllers, and the controllers control those devices, both have rich sets of data, right? We want to stream telemetry out.


There's not a lot of really good enterprise-grade things for NetFlow and telemetry and all of this.

But if we think about it, it may be under the lens of artificial intelligence now, and we can have that machine learning capability or send these terabytes of data that some of these ISPs generate daily.


We're going to throw away everything good, right? We don't care about the positive signals.

It's the negative signals and the faults and the errors and the discards and things like that.

So we throw away a lot of that data already because it's good, healthy data, unless we're trying to find healthy trends.

We cluster as much of those negative data points together as we can, and then we try to find the root cause analysis using maybe temporal data or flags or metadata, right? There's a lot of enrichment that we can do here so that the human operator can literally ask the AI, ask the agent to go find the problem or give me analysis on the problem that we're troubleshooting and come to a pretty reasonable conclusion as to what the root cause is.

So that data, that big data doesn't just sit frozen in some lake or some warehouse or something somewhere we can actually make a very good root cause analysis for one and molly you know what excites me the most especially in the networking world is the predictive possibilities i'm going to say possibilities because yes there's some capabilities with prediction right now but when they get really good at predicting based on these trends or if we feed them enough data over however many months they need them or if we can build agents that can take in multimodal data, maybe the Visio diagram of the topology, maybe the intent, maybe the actual MCP skills to connect to maybe your source of truth or your IPAM for your IP addresses.

And kind of tying back to the data portion of the conversation, since we're on data unchained, you and I were chatting a little bit when we first met about how networks generate massive amounts of data, but that remains really difficult to use, difficult to access.

Is that kind of an area that Itential is digging into, or is it more of an industry-wide problem?

I think it has been and really it exploded with the things, you know, the more capabilities the network started to carry when we get into voice and we get in over power, over Ethernet and we get over wireless and all of the hyper connected things and then maybe Internet of Things.

telemetry is trying to solve this problem, right? So you can build agents that can stream in that telemetry and help you make sense of it in human language, right? So I think that some of the barriers in the networking industry in particular around accessing the data are going to collapse.

I think that if we start to shift between humans interfacing with Originally, directly with the devices, and then more and more towards controllers, and the controllers control those devices, both have rich sets of data, right? We want to stream telemetry out.


There's not a lot of really good enterprise-grade things for NetFlow and telemetry and all of this.

But if we think about it, it may be under the lens of artificial intelligence now, and we can have that machine learning capability or send these terabytes of data that some of these ISPs generate daily.


We're going to throw away everything good, right? We don't care about the positive signals.

It's the negative signals and the faults and the errors and the discards and things like that.

So we throw away a lot of that data already because it's good, healthy data, unless we're trying to find healthy trends.

We cluster as much of those negative data points together as we can, and then we try to find the root cause analysis using maybe temporal data or flags or metadata, right? There's a lot of enrichment that we can do here so that the human operator can literally ask the AI, ask the agent to go find the problem or give me analysis on the problem that we're troubleshooting and come to a pretty reasonable conclusion as to what the root cause is.

So that data, that big data doesn't just sit frozen in some lake or some warehouse or something somewhere we can actually make a very good root cause analysis for one and molly you know what excites me the most especially in the networking world is the predictive possibilities i'm going to say possibilities because yes there's some capabilities with prediction right now but when they get really good at predicting based on these trends or if we feed them enough data over however many months they need them or if we can build agents that can take in multimodal data, maybe the Visio diagram of the topology, maybe the intent, maybe the actual MCP skills to connect to maybe your source of truth or your IPAM for your IP addresses.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.