Chris Pirillo

00:00 - 01:40

Hello, world. That's my little programming.

I'm programmed to say that. I have been since the days of the Commodore 64.

Basic. That's as far as I got with programming.

It's nice to see everybody out there virtually. I can't don't worry.

I can't the only person I can see right now is, is our guest, Gerardo Datta, field CTO for CatchPoint, and he's he's right here. And, we're both here.

It's amazing. Yep.

And happy to be hosting this conversation on behalf of, the new stack. I hope everybody has a tasty beverage out there, and you're wearing your professional pajama bottoms in watching wherever you happen to be watching somewhere on this planet unless, of course, we are being simulcast to Mars.

I don't think we are. I don't think we've made it that far yet.

But, I'm very happy, to, to be doing this today if only so that I can learn a thing or two. And speaking of learning a thing or two, we've got, you know, a few questions we wanna be asking.

But if you have a question that comes to mind in the course of this conversation, feel free to ask it. We're gonna stack those questions towards the end of of this particular broadcast.

And this broadcast, for whatever it's worth, will also be available as a download, so you'll be able to watch it after the fact, just so you know. So thank you once again, and, let's go ahead and get started because I understand I've understand boy.

Hang on. Let me have another sip of coffee.

I understand we have an amazing set of visuals to go along with the questions today, which is great for me because I happen to be a visual learner. And, Gerardo, it's very nice to, to meet you.

Gerardo Dada

01:40 - 01:52

It's a pleasure, Chris. Good to see you again.

We've been talking about, the old times and, how long we've been spending in technology business and how much has changed in the last twenty plus years. So it's great to be here with you.

Thank you.

Chris Pirillo

01:52 - 02:16

Yeah. We we we would have, I think, lightly crossed paths digitally when you were here in the Seattle area.

And, yeah, that was I we could spend this whole hour reminiscing of the way technology was, but the way it is today is a perfect, segue to, what you're doing at Catchpoint. So you describe yourself as a field CTO.

Is is there a specific reason why?

Gerardo Dada

02:16 - 02:53

Well, so the thing is, I also am responsible for marketing, but nobody wants to hear from a marketer. Right? But I'm I'm not like like you, I I started with a Commodore sixty four back in the day, learn programming, data assembly language programming, databases, networks, and so so being very technical.

So my job as a marketer is to explain technology and to interface with the world and explain how technology fits in today's world. So let's say Field CTO is just a claim to be say, like, hey.

Please give me a a little bit of the benefit of the doubt since I'm a bit more technical than the average marketer. And and thus, you know, I'm trying to add value to to a conversation.

Right?

Chris Pirillo

02:53 - 03:00

Okay. I just I I envision you somehow in the middle of a field, like, you know, like a cornfield.

Gerardo Dada

03:00 - 03:25

Well, it's it's the the field means that, travel around and give presentations to customers and work with a lot of CIOs and CTOs. You know, we're talking about Microsoft.

That was part of the executive briefing center, and we were talking about things like mobility back in 02/2005 in the early days of mobile devices and stuff like that. Yeah.

So that's why in in the field, talking to actual customers, which is a very enjoyable part of my job.

Chris Pirillo

03:25 - 03:36

Well, I mean, we've made it pretty convenient for, for this conversation to happen virtually, so no one really has to travel anywhere. Sorry if you love traveling.

You just had to travel to a computer screen. Right?

Gerardo Dada

03:36 - 03:37

Exactly.

Chris Pirillo

03:37 - 03:47

So let's go ahead and and get the ball rolling here. How has the shift to cloud computing and hybrid IT infrastructure reshaped the rule of monitoring in IT operations?

Gerardo Dada

03:47 - 07:47

Talking about reminiscing, when I moved back from Seattle a short while after that, I joined a little company called Rackspace in San Antonio. And they were launching this thing called the cloud.

They actually talked to a recruiter at Amazon. They told me they were launching this thing called cloud computing, but a recruiter could not even explain it properly.

I mean, it's it's crazy. Right? It's only fifteen years ago, less than that, that we that the cloud became a thing.

We had a cloud university to teach people what the cloud is and how you could use it. I remember doing demos to people.

Let me create a virtual server. Five seconds later, you have a server up and running, and it was mind boggling.

But now everything is on cloud. Right? Like, it's like today, you don't say saying cloud computing is like saying I have an electronic computer or is there any other kind? Everything is is cloud connected.

And the thing is for every IT department, whether you want it or not, everything is cloud centric today. Right? So, let me show you a slide that you do talking about visuals.

Applications back in the day when when I started working for a company, you have a a physical, closet where you have your servers, you have your chain server, you have your file server, you have your networks. Everything was in the building.

Right? Your even you might have Siebel running on premises. Right? You have your own, all the systems were in the building.

Now the opposite is true. Employees were all distributed.

Everything is remote. You might have in this case, you have an identity server connected remotely.

You have remote databases. You have multiple clouds.

In in fact, good cloud applications that that everybody knows nowadays, you need to be multiple availability zones, redundant services in different, locations, clusters around the planet. Right? So it's it's crazy.

So if you think about, for example, cnn. com is one of the pages that I I like showing because we have a a little tool called web page test, which you can use to run a waterfall on any website.

So just testing the website cnn. com, just the homepage flow.

This doesn't include anything happens in the back. Right? Like, all database calls or where they get information or the content management system, 625 different requests.

So it gets fonts from from one site. It gets an advertising tag from another one, Maybe the video from Vidyard or some other site, the graphics from the CDN, the JavaScript from all over the place.

So imagine for loading a page that you expect loading three seconds, six hundred different things need to load in in that time frame from six maybe not 600 locations, but dozens and dozens of different locations working together just to serve a a page. And so the point here is is that not only everything is Internet centric, but the Internet's far more complex and fragile than we like to to think.

So the way this changes for IT is is that now you you know, in the past, you needed to have a tool to monitor your servers, web server, application server, and database. Then you start monitoring your your security and people build Knox.

Now you need to think about how am I gonna monitor all these dependencies. Right? So another visual that I think you liked a lot is is this one.

Right? So in a new banking a banking system. Right.

So if you're a bank, you have things like look at the top. Payment processing.

Anybody trying to make a send money via Venmo. Any, you know, corporate companies doing swift, payroll systems, digital banking, people on mobile devices, multiple servers, your branches, your ATMs, your people working different offices, all your IT systems, your CRM.

So all this is is dramatic complexity. And to make it even more difficult, like, just think about, I don't know, let's say your your Venmo or your Swift.

Each one of those little lines has a dependency on DNS. As the joke says, it's always.

Chris Pirillo

07:47 - 07:48

It's always DNS.

Gerardo Dada

07:48 - 08:27

DNS is always a problem. And you depend on the ISP, and you depend on BGP, which is the routing protocol for the Internet.

So a ton of different things have to happen for this banking system to connect to that, API and send the information back. So so if you're in IT, one, you need to be aware of this.

And second, you need to be on top of things, and you need to be able to respond to anything that happens because the bank cannot operate unless all these things here are working well. And and that's a fundamental shift from from the way that IT has been thinking for the last, I don't know, thirty years plus.

Chris Pirillo

08:27 - 08:42

It's incredible how, interconnected these things are. I mean, if I go to a website and it's not working, I'm I'm I wanna blame the website, but it may not be the website that's not working.

It's something that it's connecting to, right, to your point.

Gerardo Dada

08:42 - 09:45

Correct. And and in in many cases, those things are dependencies that are not your responsibility.

Right? You might have a CDN. Well well, the CDN is down.

It's not my fault. What am I gonna do? Well, if you run an ecommerce website, so if you're let's say, you're the CTO or the VP of ecommerce for target.

com or Best Buy, the the CEO or the the company doesn't care if it's your responsibility or if it's a third party CDN. And they say, like, if the CDN is down, it's like, go find another CDN.

That's why most companies, of that size have multiple CDNs. We were just talking to to IBM precisely about, with n s one about traffic steering during the Super Bowl.

So we are feeding information to n s one so they can make a real time decisions about, a CDN saturation. They were actually working with multiple, CDNs so that when they saw that one of them was getting near capacity or maybe it was getting expensive, they could actually switch traffic to others or change the balance dynamically as as the game progressed.

Wow.

Chris Pirillo

09:45 - 09:59

It's I tell you, you know, It it this kinda leads into the next question though. You know, how do third party dependencies like DNS, CDNs, APIs, how do they introduce blind spots in traditional monitoring approaches?

Gerardo Dada

09:59 - 13:13

Well, so so the thing is, you know, it's it's seems like it's every every day you pick pick up the technical technology news, and you'll see this company is down or this website is down. Sometimes it's even more impactful when you see, I don't know, Southwest canceling or the flights for a day.

Right? Three years ago, we had, Facebook, Instagram, and WhatsApp. Everything went down for about eight hours.

And so the first thing to think about is that it's it's not if it's gonna happen to you or not. Right? Like, if you you might say, like, hey.

I run-in tight shape. I have good systems.

Well, it it happens to everybody. Right? It happens to Amazon.

It happens to Microsoft. It happens to Google.

We see incidents all over the place. And so the question is not and and those those companies have tremendous technology, tremendous quality of people, processes, etcetera.

So the question is not if if it's gonna happen to you. It's just is that when is it gonna happen? Are you gonna have the right systems in place? Are you gonna catch things quickly? And and so you you don't wanna be in the news.

Yeah. Right? We you you wanna make sure you're you're first to know when something happens.

That's that's critical, right, for for these systems. So the the thing that happens in monitoring let me show you another visual, is that we all this collection of technologies that are relatively new and people need to pay attention to is what we call the Internet stack.

So for years, we've talked about, you know, there's a technology stack, like, there's the seven layers of OSI for networking, and there's traditional server, like, we were talking about web server, app server, database, and there is the application stack and, you know, there's a lot of dependency mapping nowadays. But now companies need to pay attention to this Internet stack, which is by the way, it's just a it's not a a, an authoritative like, this is the entire list of things.

It's just an example of things that you need to be aware of from SaaS applications down to Internet core, to protocols, to networks that you need you wanna make sure that you are monitoring, you're paying attention to. Right? So all these things in the Internet stack.

So the incident that happened with Facebook three years ago is because there was a change in BGP. So BGP is something that most people don't know about.

It's called, border gateway protocol. It's the way that networks, send packets around the world.

So, the BGP is like the ZIP code when you're shipping something in The United States or in other parts of the world. If if your if your ZIP code is broken, and then things you don't get stuff.

Right? The ZIP code for a company like Meta, they have their own ASN and their own ZIP codes. If if the BGP routing is broken, they don't get any traffic.

So you're not able to access email. You don't access, your website.

APIs cannot communicate with you. Employees cannot get into a data center, right, because their key cards were not able to authenticate.

So it's one of those things that, you know, it didn't seem that important until it happened. And the company lost tens or maybe even hundreds of millions of dollars in revenue because they were not paying attention to BGP.

Right? And and there's not very many companies that can provide BGP information in real time. Okay.

‍

Chris Pirillo

13:13 - 13:28

That's I think that's the linchpin right there. So then what role does synthetic monitoring play in IPM, and how does it differ when performed from cloud data centers versus user a diverse user like vantage points?

Gerardo Dada

13:28 - 16:43

Well, so the the first thing is to understand the value of synthetic. Synthetic monitoring is someone like a secret shopper.

Right? Like, imagine you send a secret shopper to a store and they record the experience and they come back and say, you know what? The store was closed. Or they say, like, look.

Everything was fine, but the employee was rude. Right? You get a lot of real world information.

And they send in a secret shop where you can do that whenever you want. You don't have to wait, and sit and wait until you actually get feedback from customers.

Right? So that is that's kind of how we explain synthetic in in the most simple way. Part of the value of doing that is you can be proactive.

So most systems today, they have an error threshold, like, typically, let's say, 25%, error is that's when you trigger an alert. And as you know, in monitoring, everybody hates alerts.

Right? Like, alerts people get too many alerts, and they tend to ignore them. So let me show you an another graphic that I created that that shows, like, in typically, you have, like, change.

Like, it's in red here on time. And then that change creates problems, and then customers notice a problem.

Right? So, okay, something's wrong. I cannot, perform whatever imagine you're a bank.

I cannot perform the deposit I'm trying to make or I cannot send money. And then so customers are frustrated.

And then they when they're frustrated, they are you know, they might actually complain on social media. They might tell their friends.

They might think, hey. I I need to change to another bank.

And then, you know, some of them, as we know, in some cultures is as low as 4%. I think Japan is one of the cultures with lowest.

In others, it's 15% of people who complain. So you complain, you escalate.

Okay. Now you have the error threshold either because of the system of people.

And so, okay, you're detected. IT starts working in it.

Okay. We fix it, and now you have, like, the the resolution.

Right? And that mountain, which is say savings actually right now is a problem. Sorry.

Wrong wrong ins, label. But now imagine you're actually testing this proactively.

Every five minutes, you're sending your secret shopper to simulate that transaction on the bank from multiple locations. Right? So you you proactively you detect the issue as soon as it happens.

The detection is much faster. You have the ability to fix it if you can get to the root cause, and then you solve it.

So your mountain of incident is that little blue hill rather than the gray mountain. So all the gray stuff is all the impact that you're saving.

Right? So that that is the impact of proactive monitoring, of being or using synthetics. Now a lot of companies use synthetics.

They're they're they're all over the place. What what's really important is that a lot of synthetic technologies, deploy their agents that do the secret shopping in the cloud.

Right? So you're in Seattle. There's a data center there from Amazon.

You deploy an agent there, and from the cloud, you're testing a website or a mobile transaction. Well, the thing is, as you and I know, a data center in in in the cloud has a tremendous capacity, networking, throughput, different servers, etcetera.

Very different experience than what I get or different network capability than what I have in my phone connected to a a wireless network that might or might not work properly. Right?

Chris Pirillo

16:43 - 16:44

Just a bit.

Gerardo Dada

16:44 - 19:21

So your customers are not in the cloud, and you need to know what is the real world experience is the first thing. And the second thing is that experiences change by region.

Right? So we were talking about the, you know, we're using multiple CDNs. Many of our customers, we sell to many retailers, they actually try and test and benchmark their CDNs in different parts of the world.

We publish a report on on Gen AI tools. Right? So if you're everybody's building AI, everything.

Right? So you're seeing Copilot or Gemini or whatever via an API. Your system now rely on that API.

Well, it turns out that the response time of those APIs from Seattle where you are, from Austin where I'm I am, or from your your employees in in in Europe or, you know, other parts of the world, it changes dramatically, not only by region, but by Verner, Irvine, by region and over time. So it is important measure from where it matters.

That's what we say. You need to monitor from from the locations that matter.

If you're a luxury retailer, right, so, if you're a Porsche, they are a manufacturer, Louis Vuitton, Likely, China is is going to be your most important, market. With a great firewall of China, and most companies or I would say, I think I don't think of any any big company that actually does synthetic monitoring from inside China.

We have a hundred different agents in China. So we can tell you what is the experience of somebody in Shanghai versus somebody that is in another city outside.

I don't know many cities in China. But, in in a somebody in in a mobile device versus somebody that is connected through a different ISP, maybe China Mobile's Mobilecom, versus somebody connected through Alibaba, may versus maybe even connecting to a cloud node inside, one of the cloud Tencent cloud in in over there, and what is the connectivity back to The United States or where your service are.

So being able to understand the impact of the systems is critical, especially, you know, even today on security perspective. Most companies are use implementing SASE.

Right? Zero trust network access and different remote access technologies. You need to understand what is the impact to your users of those technologies in each region.

Because it it might work fine if you're a Seattle based company in Seattle, but if if if, you know, people in Austin, I connect through Spectrum. I used to have AT and T before.

I there there are very significant difference between the two. And there are things that change all the time, like routing protocol, size p performance, the configuration of those SaaS providers, etcetera.

So being able to react from those geographies and be aware of what's really happening in the real world is is critical.

Chris Pirillo

19:21 - 19:31

So then what unique metrics or insights does IPM provide that are crucial for managing the, if you will, the what you described, the Internet stack?

Gerardo Dada

19:31 - 22:42

Yeah. So so the the the one thing that is really critical here is understanding what is the actual experience.

Because for for years, we've been monitoring systems at the system level, obviously. Right? So we, I I used to sell database performance monitoring technologies, when I was at SolarWinds, and we talked about wait time analytics inside of SQL Server.

Right? And we can tell you, like, your SQL queries are running this fast or there's this much wait time in your databases. And and so you can also tell on the cloud monitoring space, when I was at Rackspace, you could say, like, look.

This is the CPU utilization of your cloud servers, or this is the memory thresholds that you're reaching if you have a VMware cluster, with VROPS or where wherever else you're using. Now that's useful information.

I'm not saying it is not. But what really matters is what is the actual experience you're delivering to your users.

Right? If you're a bank, you care about the mobile user trying to complete a transaction as well as the teller that is in the bank trying to serve customers. Like, it's it's almost a joke when every time you rent a car, you go to the counter, and there's always a line.

And they always say, sorry. My computer is slow.

Right? So that impacts the business and and impacts profitability and impacts loyalty and impacts trust. So what really matters at the end of day is the experience of your application.

What is experience to the either internal user, external user customer, or even a system, an an API? Right? Like, today, with Internet of things and systems communicating, through APIs and gen AI, the experience you're delivering to another system or to another user is critical. So as as an example, let me show you, another slide.

We have this thing that we developed called the experience score. So the experience score you say, look, we we it's it's not that we don't care about what is the CPU performance, which is great, and the database wait time's important, but I wanna know how is my team in Dallas able to produce.

Right? Imagine you're out in our lawyer office. I have we have office in Dallas, New York, Chicago, etcetera.

I wanna make sure those teams can actually do their job. And you can actually see experience scores by geography, by team, by individual.

We deploy agents inside, you know, and use of laptops or desktops. Which you which is we monitor the actual computer and their experience.

And and so instead of saying, look, CPU to the stage is going from 75 to 35. And as an IT guy, it's like, do I care? Do I need to do anything? Well, if you look at this dashboard, look, you need to pay attention to London.

There's something going on in London where the experience is bad, and your users are either complaining or frustrated or not productive or all of the above. So that's where you need to pay attention to.

And then from there, then you can be you can you're able to look at, okay, what is costing that experience? Is it the network? Is it the application? Is it an API? Is it a provider, etcetera? And then you can quickly solve the problem. So it actually points you to where you need to go versus just looking at a bunch of system metrics.

Chris Pirillo

22:42 - 22:53

So then how do you IPM and APM work together to create a a more comprehensive monitoring strategy for resilient digital experiences?

Gerardo Dada

22:53 - 24:57

Yeah. It's a it's a great question because, you know, IPM is, part of observability, and a lot of people, understand that that, you know, they have APM and a lot of companies have NPM.

And so what's important is to understand what is the information you're getting from those systems. Right? So NPM is gonna tell you about your local networks and traffic.

APM is gonna tell you mostly about the application. Right? And and so even the synthetics they do are great for monitoring the performance of new codes.

We're so much like a QA process. It's okay to do that from the cloud.

But the focus of application performance monitoring is is really application logs, code traces, and infrastructure metrics. Right? Or and also event.

Like, now there there's this acronym called NELT. IPM looks at experiences, what is the health of the global Internet, what is the experience of your employees.

So the the idea is that now you have NPM, APM, and IPM, all three of them working together. So we we put a special care into exporting our data, making it available to open telemetry.

We can consume data from from your your agents' inside code and show traces as part of the IPM, but also export our data to other systems or, you know, big companies use things like Databricks or other data systems to collect and monitor systems. But the idea is that the majority of large, mature companies are using APM and IPM together to get the best visibility they they need into their systems.

So you can see both the end user experience, the Internet visibility, and also your internal visibility into your your technology and your application, your code. It's like in your car.

You have a dashboard that will tell you what is the RPM and and the the gas capacity of your tank and, some other systems that are your system metrics. And then you have your GPS dashboards.

It's gonna tell you where are you going, when are you gonna get there. There are incidents that you might wanna take a different location, a different route to get there.

Chris Pirillo

24:57 - 25:15

Yeah. Now speaking then of kind of working together to create a better experience as the Internet becomes more integral to business operations, what steps can companies take to align IT monitoring strategies with business goals, like customer satisfaction and, operational resilience?

Gerardo Dada

25:15 - 28:26

Well, so so the first thing that I I think it's it's very interesting for a lot of these businesses is to think about the idea that audio interactions are digital today. Right? Mhmm.

So, we let me show you a graphic that shows, like, what we call a dock. So people have a NOC, a network operation center, and they create now a SOC, security operation center.

I think company need to build a dock, a digital operation center. So this basically tells you how is our business working as a digital system.

Right? If you're a bank, like my mobile applications, my ATMs, my internal people, my traders, my people in my offices, HR department, etcetera. You wanna make sure that all the digital systems that are essential for your business to work are operating web.

Right? So you wanna see that dashboard with all grades. And we see this at at some of our customers.

It's it's I I don't know that I can mention the names, but we walked into some of these operation centers that look like the the NASH emission control, and their main screen at the center is a cash flow system like this that shows you what is green and what is red and what needs attention. And from there, you can look, okay, let's look at the network and let's look at databases, let's look at applications, etcetera.

So that's that's the first thing I think it's important for customers. The second thing is is to think about, let me show you something else, is to think about different metrics that you should be thinking about.

So there is a a traditionally, most businesses operate on on the this concept of SLAs. Right? So you have an SLA for for different systems.

And and nowadays, you wanna those SLAs are actually giving you a performance of the of specific parts of the system itself. Right? So it's not showing you, the actual user experience.

And and so the business cares about the user experience. IT seems to care about system metrics.

Right? So so now imagine a system where you can actually adopt, user metrics as your end goals. So that's what we call excellence, experience level objectives.

So so then imagine if your SLA, instead of saying, you know, my my system needs to be able to to operate at, 70% CPU utilization again, you say, like, I wanna make sure that my users in any of the top 20 cities in The United States can complete the transaction within ten seconds. I wanna make sure that whenever everywhere when we have a rental car location, our agents are able to pull out our reservation and submit it and get responded within four seconds.

Right? So that's an experience level objective. And that basically puts the end user at the front of of the system.

And that means that what IT is striving for, which is I wanna make sure that the IT systems in this rental car company are supporting the experience that we deliver to customers, supporting the business. That's the same objective from the business and from IT.

And any improvements you make to the IT systems will be improvements that will apply, that would you will be able to show us how you're meeting or exceeding those experience level objectives.

Chris Pirillo

28:26 - 28:45

Okay. So then if you can break it down for me, what's the primary difference between digital digital experience monitoring or DEM and Internet performance monitoring or IPM? And how do those tools complement application performance monitoring or APM?

Gerardo Dada

28:45 - 30:58

Yeah. So digital experience monitoring is kind of a there's there's different ways to think about it.

Right? But, you know, if you think it like, most APM systems offer some level of digital experience monitor because it's so important. Gartner Research published for the first time the the magic quadrant for digital experience monitoring just a few months ago, recognizing that it's so important.

But in the same report, they said, well, there's also this thing called Internet performance monitoring, which is basically well, let me step back. Digital experience monitoring is the ability to use things like synthetic monitoring or wrong technologies of other systems to understand that experience you're delivering to customers.

Internet performance monitoring expands that definition to say, yes, but you need to monitor for what your users are. And second, you need to understand all the impact of anything in the Internet that is impacting that user experience.

So it's not only useful to say, hey. Your website is down.

Your customers in in Seattle cannot access your website. You wanna know why.

You wanna know what is causing that problem and and how you can solve it. Right? So it's not enough to say your API has a higher response time than what we want.

You wanna know is that a routing issue? Is that a response time issue? Is there something inside the database that we need to go fix? Like, that that is the the main difference between DEM and IPM. There are very few companies that can do IPM properly precisely because of the need to have this geographical view.

So let me show you. For example, we we have a a a network of intelligent agents around the world.

We have, almost 3,000 intelligent agents. So I was talking about hundred of them.

In China, we support over a hundred different countries. We support, we have over 1,500 different BGP peers, different locations, different providers.

And that's what makes, becoming an IPM provider really difficult is that you need to have all this global presence, not only geographically distributed, but also inside the cloud, inside the backbone of the Internet, in last miles. We can even monitor Starling performance.

Right? So we have nodes that that are used in Starling so that we can show you what is the satellite connectivity. We haven't done tomorrow's yet, but who knows? It might be, it looks.

Chris Pirillo

30:58 - 31:12

Next week. Now can you provide examples of real world incidents where IPM helped resolve performance or resilience issues faster than traditional tools?

Gerardo Dada

31:12 - 33:50

Yeah. So let me start with one we call workforce experience.

We we were dealing with a, let me pick, there's a technology company that sells a lot of computers. And their Salesforce was complaining all the time, hey.

My I'm connecting to salesforce. com or I'm using different applications or entering my orders, and and I'm frustrated.

Like, Gartner Research says that there's a tremendous amount of frustration. They call it digital friction.

It's something like eighty percent of employees suffer from digital friction, significant digital friction on a weekly basis. So it's typical that they were getting complaints and IT was saying it must be used at all.

So, the there was impact on on the trust of the employees. There was impact on their productivity.

There was impact also on on, you know, the trust between IT and the users and and the productivity of of those salespeople. And they've been trying to solve this problem for a long time.

They deployed CatchPoint, and they found that there there were some incidents in in the applications and the ISPs that we're using. Mhmm.

And they were able to catch those incidents in within minutes and solve it. And now now you have thousands of sellers who are normally happier at the job, but they actually can save ten minutes a day.

Now multiply ten minutes a day of saving time, times 5,000 employees times three hundred days at a year. It's significant savings from a workforce experience.

We're also working with with Sony PlayStation. The one of their biggest markets is in The Middle East, and they were having an incident, and and their customers were were complaining, and they could not figure out why why were customers complaining.

They said, we've tried every monitoring tool. Everything looks green on my side, and yet customers are complaining.

Right? That's a that's a great indication that you need IPM. All my systems are green.

All my monitoring tools say that everything's fine and get my entire set of customers as frustrated in this particular region of the world. They install Cashpoint, and within minutes, they we showed them that there was a component of their application that they have hard coded to route traffic to Germany.

So instead of using the local CDN or the local servers, traffic was going all over the world to to get, to what it needs to go. And in in in in gaming, latency is very critical.

Right? So you wanna make sure that you you're aware of those systems. And and from that point, Sony has been now using CashPoint extensively, and their customers are happier.

Forced research shows that the companies using IP and have fewer incidents, and they save money, and they deliver better experience to users. We're happy to share that that report with with anybody who's who cares.

It's on our website.

Chris Pirillo

33:50 - 34:04

So, you know, you're making the case, you know, clearly, but we're all being asked to cut costs and consolidate tools. How can we convince leadership we need an IPM platform added to the tech stack?

Gerardo Dada

34:04 - 35:05

Well, the first thing is is that we we do a lot of consolidation. Right? So we we there's a lot of companies who are using legacy tools, and we have a program to help migrate scripts and and up upgrade their their monitoring.

In many cases, margin with customer can be cheaper than some of the alternatives. The more important cost is what is the cost for your trust with your users, and what is the cost of your, of the war rooms.

Right? So if you are money if you're not doing proper monitoring, every time there's an incident, you typically call a war room, and that costs a ton of time of your best people who need to pay attention to, hey, what's broken. And and it's like triage.

Right? Going one by one of the systems to try to find the problem to kneel in a haystack. That costs a lot of time and and effort from an IT department as well.

So that's the first part. And and the second point is is that when you can find the problem quickly and deliver better experience, it's better for the business in general.

‍

Chris Pirillo

35:05 - 35:16

So everyone's talking about AI. How can AI driven insights such as those provided by tools like Internet sonar enhance root cause analysis and resolution times?

Gerardo Dada

35:16 - 38:09

Yeah. So that's that's, always well, we've been doing AI for maybe over ten years.

So let me show you. We have this thing called, Sonar, which, we have a free version on the website.

Everybody can go to cashflow. com/outages.

And what I'm showing right now is, the live outages. We have a weather service, Asana, chief flight, Salesforce, and Hulu, all done in different places.

So we run billions of tests a day for hundreds and hundreds of services separate from the tests that our customers want us to run for them just to get a general health of the Internet. So we use AI because we don't wanna alert people for every single little thing that is happening.

Right? So you you wanna use AI to say, like, is this actually a trend? Is it happening multiple times? Is that one failure enough to create an alert, or do we wait until we have more? So we correlate and and consolidated all this information. Let me show you something else.

There's this other chart that I wanna show you that is is called stack map. So stack map uses correlation and trends analysis to actually detect what is the right, when what is the the cost the root cause of a problem.

So in this case, what I'm showing right now is a application dependency map that shows not not the traditional dependency map that Joaquin is focused on the system, which is code to code. This is actually from the end user, origin servers, DNS, cloud providers, applications, APIs, all the way to the back end to database and to servers using tracing.

So using AI, first, we create this map. Second, we normalize the data.

We, identify trends that require attention, and then we identify what is the root cause that is that is that requires where where you need to pay attention to solve a problem. Right? So you can then drill down and and do everything you want, but this is a a good example of AI.

Yet another one that we have that I I particularly like is on the web page test technology that I was talking about. We do this thing, the the the you know, I showed the CNN example with the 600 different things.

We have AI that detects, like, we think there are these experiments that we wanna run. There are opportunities to improve the performance.

So we actually, with AI, generate experiments that simulate the basically, clone the website, make the changes that we think are gonna improve the performance, run our tests again, and we can show you, hey. We think by a deferring JavaScript, this JavaScript line or by optimizing websites, you can you can improve performance.

But let me test it. We actually test it and tell you, yes.

It did work. Or you can test the combination of steps.

Tonight, you have the certainty that the changes you're recommending actually will make a difference for the end users.

Chris Pirillo

38:09 - 38:53

Well, you know, this is I certainly, I I love having these types of conversations because I get a chance to learn a little more about how they say you never wanna know how the sausage is made, but how the sausage is made. And so, you know, I certainly appreciate the the insight, and I know, people who've tuned in appreciate what CatchPoint, you know, is or can provide for for them.

So we're gonna start winding down here, but I do wanna ask, like, one last question here, unless there's any that that have come in from from the viewers that I I feel has not yet been covered. What should organizations look for in an IPM platform to ensure robust visibility and performance management?

Gerardo Dada

38:53 - 39:57

Well, they should look at three things. One is that the you need to monitor from work where it matters.

Right? Where your users are. Again, users could be employees, customers, or digital systems consuming API.

Second, you need to monitor what matters. What matters is things from BGP to your CDN to ISPs to APIs to your own internal matters.

By the way, there's a question about monitoring cloud services like AWS. The the the the sonar graph that I was showing you actually monitors those cloud services in real time so that you can be aware when, I don't know, AWS compute east is down, for example.

So that's monitoring what matters. You can monitor inside, peer to peer connectivity between two network clouds.

You can monitor from your systems to the cloud. All all those kind of things.

So it might turn on what matters. And the last one is get to the answer faster.

Right? You solve the tools and the metrics and the right systems so that whenever you detect an incident, you can react quickly and and get the answer meaning, solve the problem ideally before your users are impacted.

Chris Pirillo

39:57 - 40:21

You know, Gerardo, like I said, it this is it's fascinating to me. This is, like, beyond I'm the user.

I I'm the one going to the website or using the services going, it's not working. So I need I, on behalf of all the users on the world, need companies to use CatchPoint because I do not I don't wanna I don't wanna mess with downtime.

Who wants that? No one's got time for downtime. Nobody.

Gerardo Dada

40:21 - 41:07

Yeah. Every business now relies on the Internet.

Right? We we like to say the Internet relies on Cashpoint. That's why all the cloud providers, Amazon, Google, Azure, they all use Cashpoint.

The seven of the top 10 digital companies, the the top software companies in in the world use CatchPoint. Every CDN on the planet uses CatchPoint, a lot of the ISPs, every, commerce platform from SAP commerce, Salesforce commerce, Shopify, because it's it's essential.

Right? It's not it's not optional anymore. It's not this one thing that it it is it is cool and interesting.

It it is essential if your business is digital and your users are digital, which I argue pretty much every business is, then you you need to pay attention to LinkedIn performance monitoring before you show up in the news. Right?

Chris Pirillo

41:07 - 41:12

Yeah. You can't just look out the window and say, well, it's it's sunny here, so what's the problem?

Gerardo Dada

41:12 - 41:14

Yeah. Exactly.

That's good analogy.

Chris Pirillo

41:14 - 41:29

And it it is sunny in Seattle today for whatever it's worth. It's an anomaly.

I don't know. First day of spray fake spring.

I don't know exactly what's going on. Well, anyway, thanks again for your time, your expertise, and, taking the time to have this conversation with us.

Gerardo Dada

41:29 - 41:32

Thank you, Chris. It was fun.

Appreciate your time and those questions.

Chris Pirillo

41:32 - 41:36

I appreciate the visuals. I genuinely do.

Like, I I mean, I it's all in the.

Gerardo Dada

41:36 - 41:40

a visual person always. Yeah.

Lot lot of whiteboarding sessions and stuff like that.

Chris Pirillo

41:40 - 42:02

Yeah. Yeah.

Big time. Big time.

So, again, on behalf of the new stack, really appreciate you being here and doing this. Everybody, this will be available if you're watching live as a download after the fact.

If you're watching the download well, you know, you're watching the download. So, hey, it's a good thing.

Thanks again, everybody, and we'll catch you the next time.

Gerardo Dada

42:02 - 42:03

Thanks, Chris. Thanks, everybody.

Chris Pirillo

42:03 - 42:04

Thank you.

‍

‍