5 Ways Docusign Solves Network, App, and Device Issues Faster and Stopped the Firefighting with ZDX

  • Zscaler
5 Ways Docusign Solves Network, App, and Device Issues Faster and Stopped the Firefighting with ZDX

About This Webinar

The Standard Edition of Zscaler Digital Experience (ZDX) scores users' digital experiences. But as hybrid work spreads workforces across unmanaged home ISPs, diverse endpoint environments, and critical SaaS apps, the ultimate goal is turning those scores into proactive detection and instant root cause identification.
Join Docusign Senior Manager of Network Engineering and Collaboration Platform Rohit Bora and Lead Site Reliability Engineer Shoubha Chakrabarty as they share how they built an evidence-based operating model using ZDX Advanced Plus across networks, applications, and devices to stop the default network blame, eliminate 20-40-person war rooms, and pinpoint root causes in minutes, not days. All using the same client connector that keeps users secure.

In this session, you will see how to:

  1. Use AI for Accurate L1 helpdesk triage: Ask ZDX Agent to get an instant, plain-English root cause diagnosis without digging through data
  2. Deflect ticket waves: Monitor application outages, highlighting Docusign's over 90-minute head start on a vendor outage as a blueprint
  3. Pinpoint root cause instantly: Isolate slow app issues to local Wi-Fi, home ISPs, device processes, or SaaS vendors in seconds
  4. Prevent device issues proactively: Detect high CPU, device bottlenecks, and application crashes before tickets are filed

Speakers

  1. Scott Bekker Webinar Moderator Future B2B
  2. John T. Jones Sr. Product Marketing Manager, ZDX Zscaler
  3. Luis Rico Principal Product Specialist, ZDX Zscaler
  4. Rohit Bora Senior Manager, Network Engineering and Collaboration Platform, Docusign
  5. Shoubha Chakrabarty Lead Site Reliability Engineer, Docusign
Full Transcript Scott Bekker Hello, and welcome to five ways Docusign solves network app and device issues faster and stop the firefighting with zedX. Today's webinar is sponsored by Zscaler and produced by Actual Tech Media. My name is Scott Becker with Actual Tech. I'm excited to be your moderator for this webinar. So we've got a ton to cover today and, can't can't wait to get started. But first, I do wanna go over a few housekeeping details that are gonna help you get the most out of this session. We've got three tabs on the right side. There's the chat tab for your hellos, other comments, and and and that's a good spot to go if you need technical support. There's the doc section, which is the next tab that has great stuff from Zscaler, including a ZDX practitioner's guide, virtual workshop, and interactive demo, and a Siemens customer story. Then there's the q and a tab, which is where you should go to put your questions for Zscaler's, q and a. And you may have also noticed the prize icon there on the slide. We're in fact giving away an Amazon gift card worth $250 as a prize drawing at the end of our webinar. Now you must be in attendance for the duration of the live event to qualify for the prize, you've got to meet the actual Tech Media prize terms and conditions, which you can also find in the docs tab. Okay. Well, we've got a, a big crew of presenters, but I'm only gonna mention one of them, John t Jones, who's a senior product marketing manager for ZDX at CScaler. John is gonna act as our MC today, so I'm gonna turn things over to him to handle, the rest of the intros and and to get it going. John? Video Thank you. Thank you for joining us today for our our webinar, five ways that Docusign solves network app and device issues faster and stopped the firefighting with ZDX. Let's jump right into our agenda. Housekeeping, I'll cover that right now. Feel free to ask questions along the way. We will answer those, as we are able or we'll, get through them, and follow-up afterward. We'll go through some introductions. We'll go through what happens as organizations evolve and grow. Then we'll have a great conversation, with our two guest speakers here from Docusign, hearing their story and, go through, the demo about a twenty minute demo, five ways that zDx gets to root cause faster. So let's jump right in, to our speakers. Rohit Bora from Docusign, senior engineering manager for network infrastructure. Rohit will join us here along with, Shuba, Chakrabarti, who is a lead site reliability engineer at at Docusign, and they'll join us here in a moment for for that, great interview. And then Luis Rico, principal product specialist, will join me in the latter half of the webinar for a a live demo to see in action everything that, Rohit and Subha will be talking about. And my name is John Jones. I am a senior product marketing manager for Zscaler digital experience here at Zscaler, and I'm excited to be your emcee and guide today's discussion. So as as organizations evolve through increased hybrid work and, you know, growing non managed ISPs and UCaaS calls become the standard, not to mention the fact that AI app and and agent use is exploding, that causes all sorts of vulnerabilities to traditional networks as we know it. And as a result, blind spots start growing. You know, according to ESG, 74% of companies find measuring employee experience across apps challenging, especially as as silos continue to grow more tools, more unmanaged networks. That causes delayed detection. So, a global IT customer, of of Zscaler, was facing twenty four hours to detect ISP outages, and we'll talk a little bit about, what they were able to do a little later on. That means that tickets start bouncing. They come in through l one, analysts, but they lack the ability to accurately triage, and that typically then goes to a network. Oh, it has to be the network. It has to be the network, and then everyone's trying to fight that blame game. It's caused by the the silos… Siloed tools that organizations deal with. And without that single view, it's hard to get first ticket alignment, accurately. Most of the time, it's just a guess. And what's the ultimate result of this, the impact to the organization, lost productivity, and one of my favorite, stats or the… One of the more concerning stats, according to Gartner, $336,000 per hour in unplanned downtime. And as a result, like, Docusign, like many others, they turn to zero trust SaaSy architecture for security, but also for user experience. And the, one of the great stats from the enterprise management association fortunately fortunately, Zscaler does that by turning, digital experience, changing security into also a productivity edge. And, through our native client connector, all running through the same client connector that operates security, gives you great telemetry across the hop by hop, you know, journey, proactive detection, accurate l one root cause analysis, the ability to to consolidate tools and improve productivity. And what we're chatting about here today specifically is organizations that go from z d x standard to z d x advanced plus. Getting capabilities like network intelligence, being able to increase the number of probes, through cloud and web path from six to a thousand, proactive alerts from three to to a 100, and so many more things that we'll be talking about, in our conversation. And without further ado, I want to come to that conversation and hear Docusign story because their, words and experience is far more valuable than anything I could tell you. So Rohit Shuba, thank you so very much for taking the time to to have this discussion and share your story. Kicking things off, let's start with some introductions. And, Rohit, starting with you, if you wouldn't mind, giving a brief introduction of of your role, and what you do at Docusign. Sure. Thanks, John. Hello, everyone. I'm Rohit Borah, senior manager for network engineering platform at Docusign. I lead the network platform strategy, and primarily, behind the Zscaler adoption, including how we enabled zDx telemetry and incorporated that into our operational visibility, incident response, overall, you know, employee experience improvements. So been leading this platform along with overall network platform for DocuSign. Yeah. Absolutely. And then you were one of the main champions of of of this this journey. So excited to have your perspective here. Shuba, same question to you. A little bit about your your role and what your day to day looks like. Yeah. Hi. So I'm Shuba Chakurty, and I'm a lead SRE on the infrastructure side of DocuSign. So, primarily, our team, provides SRE expertise for, infrastructure, which includes network, and that's naturally one of the reasons why I partner so closely with Rohit. And, we also monitors, enterprise application systems and data pipelines. And network is at the core of everything, so naturally, I've partnered with, Rohit on the z d x implementation and the rollout. Yep. Yep. Absolutely. The network touches all things, which is also why it's the the most vulnerable thing that we try to, to remove. So, jumping in to, to to kind of starting to set the stage, Rohit. Before you had deployed zDx, when an end user had reported, you know, the the tickets are always super specific. Right? And, you know, the reality is, you know, oh, the network is slow. What did that triage work flow look like, and how many tickets or how many teams did the tickets touch before it actually landed on your desk? Right. So before zDx, the network is slow, was pretty much like, you know, a symptom. There was no diagnosis. Almost one third of the tickets that came in, had this very vague term. And this is obviously how end users actually experience the issue, but also how they're able to communicate without, you know, having granular data of where that issue lies. And so the network ticket queue basically became this landing spot for everything that we could catch up Even though the root cause analysis at the end demonstrated that a lot of these issues were actually, like, endpoint processes, you know, browser specific issues, home Wi Fi, public Wi Fi, or even in the ISP path. Right? A lot of the zscaler service edges at times cause problems or the the SaaS application in itself were, you know, flaky, but all of that was just determined in this. As you said, the network is slow. Right? And I'm… Monitoring was pretty useful. Like, we we were able to get to the root cause eventually. Sure. But it was very fragmented. Right? So we had the endpoint in itself, the network, the application signals that were coming in. All of these monitoring existed, but they lived in its separate tools, separate ownership. And so for someone like Shuba, if he had to go and, like, stitch that together, was just tremendous amount of, like, level of effort. So that that is what ended up creating, like, this network team black hole, of everything getting centralized to us, having Shuba basically, you know, correlating this. And most of it was proving, you know, as part of the investigation of, hey, the network is not broken. Right? Maybe we should be focusing our energy and time in other domains. And so this cold coordination was very expensive. We used to have like 20 to 40 people, meetings at times debating operational tickets. And these were happening at times like twice a week, right? With the war room kind of situation. So you could imagine right from the service desk that that analyze these tickets all the way to the C level folks, getting engaged the, the, the, the cost, the ineffectiveness, and also, you know, conflicts in the culture overall because nobody likes to deal with ops for long. So zDx is what, helped us get that single pane of glass. Right? So one experience view, from a user's device through the network path to the destination app. Right? We could stitch those, end to end telemetry. And the win wasn't just having, you know, these these additional graphs and correlation, but it is this data driven approach now, to tickets, which achieved, like, faster attribution, cleaner handoffs. So I'll I'll I'll stop there, but that's pretty much the before ZDX. Absolutely. And the thing that really, you know, stuck out to me in in our initial conversations was what that process looked like. Because you were spending so much of your time say whose problem it wasn't instead of actually going to whose problem it was to fix. And that 20 to 40 person war with all of those people, the amount of time, money, resources that were tied up in those conversations to a root cause. So, I think that is something that most people on this call are going to to experience. But, shifting over to to your side, Shuba, you know, from a from an SRE perspective and and support side, how did you handle… You know, we talk about the the non managed network side of things. How are you handling home versus off work net… Network complexities, and what did it take, to actually get a single ticket resolved? Yeah. So continuing on what Rohit said. Right? So it was all siloed. Right? So first of all, home, we we never had visibility. So, if a user said that there was an issue that, the user is facing and if, unfortunately, that location was the user's home, we don't… We didn't have any, visibility into that. And it would take, an SRE, and an end user support team member. Occasionally, the network, team, member to be on a joint call with the user to actually see, what's happening. So home was always a more complex situation. Office was a little bit better because most of the info was under our monitoring and we knew. But, again, as Rohit said, there was a there was a tool for the Wi Fi metrics. There was a tool for the somewhat endpoint data because it's not always one thing. You have to combine the data for the endpoint, the Wi Fi, and then the ISP itself. So, yeah, it… A single ticket would take me at least an hour to just stitch the data together. Wow. And then then just… Then the analysis is, again, another, difficult phase of it. So, yeah, every ticket that was there for this kind of situations used to take at least an hour to at least gather all the data from all the systems. Yeah. Plus the time that you and others were on that call using… On that call. Yes. End user having to be on there at the same time. While that's a great white glove experience, it's not an efficient way, to to solve problems. Scaled. Right? It was… It's not scalable. Yeah. Absolutely. And that's what we're talking about here of, you know, what you're able to do. And, passing it back to you, Rohit, you know, your team had initially, tested Advanced Plus during a a a proof of value, ended up initially going with a different addition of the platform. What were the additional business needs that you had that ended up, you know, leading to to your, Advanced Plus adoption? Yeah. I think our starting point was was something that that had, like, a grand vision. Right? But initial requirement was not just to build another up down monitor that is giving us, like, very high level signals. And so we needed some some intelligence. Right? Like, in terms of having, you know, end to end visibility for this truly global hybrid workforce. And we started this analysis when RTO was in its full swing. Right? But, you know, we truly had people distributed across, geos, across locations, home, office, and, you know, remote remote folks traveling. And so the needs were different at different levels. What we noticed was the service desk needed help in interpreting the telemetry data. And so just giving them some complex graphs, it is hard for for, you know, the the end user support teams and help desk teams to translate that. For me, the network engineering teams need were much more granular to get the hop by hop cloud path and have those granular evidence, you know, to to to pinpoint where that root cause actually, you know, maybe or, you know, head into that direction. The endpoint teams had a bunch of processes, crash related, and and in general, fleet health, requirements where they needed more specific signals. And then, obviously, as Shuba mentioned, SRE had their own, you know, local versus regional. Is this like a SaaS wide issue or is this more of, you know, contained in a specific location? And so I'll summarize this into, like, most importantly, we were looking for effective cross functional data correlation, right? So that these teams that have different needs can eventually land on the same platform and have that data correlation right in front of them. So with evolving zX, inside the advanced plus case was built, mainly on, like, what actions will we be taking with this with this upgrade and what are the problems we can solve? So the question was never like, is, you know, is the user not happy with the infra? It was mainly like, what can this platform tell us specifically so that the right action can go to the right owner, right? We avoid all this chaos that we just spoke about in terms of operational incident response. So deeper app and endpoint visibility, user facing guidance in terms of like real user monitoring, much richer incident context where we could look at the path hops, you know, and easy terms determine what's needed. And I think there are a lot of automation benefits as well with advanced plus that came in, that were in our roadmap. So we gathered broader business support, when we tied all of these together, presented, you know, what what the real operating cost benefit would be. And if we don't do this, what's the cost that is spent with these repetitive troubleshooting and cross functional, you know, escalations, or even, you know, spending time with, like, a lot of our SaaS and ISP providers. So we now can go to them with a lot more context and data. Right? So, that's… That was primarily what drove us to this upgrade to Advanced Plus. Yeah. And you touched on a lot of amazing things there, and one of the things that often happens is everyone wants to point the finger both internally and externally. You reach out to the ISP. Oh, no. It might… Must be the app. You reach out the app. Oh, it must be the ISP, and now you get to have the the affirmative data and the the evidence to to say, nope. This is actually you. Here's why. Here's what we're seeing. You also touch on a few specific capabilities, and and, Shuba, I'll pass it over to you to kinda uncover some of those. Could you walk through, in the more technical findings that proved that, you know, standard wasn't enough? And and what did Advanced Plus allow you to see that finally resolved all these persistent endpoint and application performance issues? Yeah. Sure. So, so at the end of the day, of course, zDx is one of the tools that we, as I use, very, diligently, and it's part of my day to day, working. Right? But, ultimately, we would want to give a tool to our end user support and empower them. Right? So one of the key things that we needed from the advanced plus queue was the, agent, you know, the ZDF's agent. Because as Rohit was saying, like, for me or Rohit's team, we need more granular data, when we get involved. And that we can decipher from complex graphs and things which Rohit was mentioning. But imagine 10 tickets came from a network issue from end end users. Right? And if you have a simple chat interface to just ask what issues this user was facing during this time. And it not only tells you the things that it saw as issues, but it also suggests recommendations. Yeah. Right? So that was a very big, driver for Advanced Plus because, ultimately, for ops to work, we would want the end user support team to be empowered to use the tool. Right? And the other thing was the probes. Right? So I think the the other SKU had, some… I don't exactly remember the number of probes that… But, of course, Advanced Plus, I think, is the, it gives the highest number of probes that you can do with z d x. And why probes are important is if, say, you have 20 enterprise applications which 90% of your workforce uses. Right? And if you can have probes on them, then, when they are flaky, it won't be… Is the network having problem or the application having problem? That Yeah. That is the biggest thing that we get to the root cause now within few minutes rather than investigating whether… Because initially, when we didn't have this kind of probing, and if all your users' machines are actually testing on your behalf as probes, it's very easy to say if 20 devices had an issue with this application. Yes. Right? But other applications are not having issues from that location. It's definitely the application. Yep. So so that, like, that immediate, root cause, analysis just by looking at where the issue is was was… And and for that, you need a higher number of probes as you have higher number of applications, which are very critical to the business. Yeah. And and what you're specifically referencing there. So in standard, you get six probes, but we know that, or organizations, especially of your size, but even even a smaller organization is is needing to monitor more than six different applications network, things of that nature. So, with Advanced Plus, you get to expand that to a thousand, and and that really gives great coverage across the organization. Right. Right. So that was the… Also the number of alerts that you can create. I don't… I think I I I think the standard one, has… I don't… Three. Three. Yeah. Yeah. So you…… 100. Yeah. So you… Yeah. Because there are different categories of alerts. One and one is network alerts. One is application alerts. One is RAM alerts. Yeah. So so you definitely cannot, cover all of those with three alerts. Yeah. Absolutely. And and the ability to then expand that proactive detection to the organization. Yes. And we'll get to see a number of the things that you referenced in in our demo with with Lewis here in a bit. Wanted to to look, you know, to some specific examples. So you talked about the l one analyst, but if you could kind of continue a little more and give some specific examples where where ZedX Advanced Plus really stood out and, kind of some of those use cases that it really proved it. And from our conversations, few things come to mind. The the beating the public static, status page of, a major SaaS provider, the impact to your VIP user support. And you've already touched on a little bit of that AI diagnosis that the agent did. So if you could cover those other two, that'd be great. Sure. Sure. So, just just to give a setup how we work with, the z d x alerts is we let z d x do the thinking Why this is an alert. And then we, ingest that alert to our standard monitoring tool that we have. And then based on the location from where it triggered, if it's, say, San Francisco. Right? So we have… Based on our office locations, we have local office channels where individual users who are from that location are there. So what happens is if a alert triggered for a SFO location, we immediately… We ingest it from a z d x and we use our… We have a a zero touch, incident response kind of a model. So what happens is it immediately routes it to the local channel for that location. Yep. What it does is it tells the end users that it's not only them that they're having issues with a particular application, but it's a broader thing. And then we kind of follow-up on that thread that, okay. This is what's happening. Because, typically, if people are working from their desk and they're having an issue with an application, they wouldn't know if others are having to. And it's it's not in their control at that point. But it's important for us to let our end users, who are kind of our customers, know, like, that this is a bigger issue. It is not related to their machine or something. So that's one of the biggest wins that we could immediately route the information where it needs to go. And also, we are not spamming the entire company when it's happening only with a particular ISP in a particular location. We are just informing that local team. So that's one of the things that we could achieve with, the z d accelerating because of the fact that you can filter on the z scale locations Yes. That you create for each office. So that's one. And the, second thing, yes. There was a, I mean, a major, collaboration tool had an outage, and they… And we kind of detected it with Zscaler because we had those probes probing that application. We knew it at least an hour and forty minutes before they announced it. Yep. So… And and as… By the same, flow, it went to the end users directly, because at that time, it's flakies and and and intermittent. But we let them know that, hey. It's something bigger is going on with this application. So it's… We know about this. Please don't send in tickets to fraud system…. Tickets. It's just that because they would be worried. Is is it my endpoint that slow, or is it Yep. The network that's slow, or what's going on? So just to clear the confusion and… So that everybody is on the same page, it… It's, it was a great win. And also for the agents. So one of our, executives, at home was having issues. Right? So as I said, the first thing we started with the agent, and we asked that, hey. For this user, when the user is at home, they're experiencing some degradation on the network What are what are the issues that you see? Yeah. So first, clearly, it said that the ISP router That the user had Yeah. Was was having degradations, which was later confirmed by by the ISP team themselves. Like, you know, the commercial ISP is they came and fixed the router. And the other thing that it detected is that the house has two networks, which we normally do. One is the ISP Wi Fi that already comes with the router. Right? And one is the mesh or the, repeaters that you buy and add it in your house. So it was clearly telling us that use the mesh network because it's more clearer Yeah. Than the, ISP's Wi Fi. So… And immediately, the user could see, some, difference in the performance. So, yeah, those are, like, very, very, I would say, I… Previously, I couldn't have done that. We had no visibility into a user's home network. Right? So so… Incredibly important to keep those VIP users productive because they're also the ones that have the biggest influence, to, you know, cause issues to architecture and and influence the kind of rollback that that would be detrimental to an organization. Right. And also just just because… I mean, when when when the user already has the infra to have a better network Yeah. At least letting the user know which one is the better one instead. Absolutely. Absolutely. And And I love the example of proactive detection and alerting of of of the users. But… And we've talked a lot about network and applications, but you've also done a lot with, you know, endpoint, device. What's that visibility allowed you to do? Oh oh, yes. So first of all, so with Advanced Plus, we get this crash reports. Right? And, for application crashes. And, of course, what happens is, what we have done, now, we took the top three, I would say, offending applications that would crash very frequently and, like, in thousands In number of times. Right? And we would take that information and give it to our, CPE team, which is our, client platform engineering team. Yep. And once they have the signal, they know what to do, and they they immediately would work on it and deploy the fixes. Things like… Even, like, updating, some, drivers. Right? Sure. Also, one interesting thing is one, this video conferencing tools. Right? They… Even even people have issues, their jittery and, you know, maybe… Sometimes it's it's as simple as that it has a a a broken installation. Right? Some update, was going on and it couldn't complete. And it's in a a weird state where it's not functioning the way it's supposed to. Even those… Yeah. And consuming high, CPU. Right? All it took is to get that report, like, how many of such devices are there where this particular application is causing the problem, and just reinstall that application, and it it fixed it. So things like that have been great for the feet, fleet health, and we want to do this, more in a automated way. As of now, we are consuming the reports, which is also easy, but, we would like to automate this further. Yeah. And and and just amazing. We'll see some of that in the device health dashboard, in our demo in a little bit. Wanna shift gears a little bit. You know, amazing examples of of the use cases of of what you've been able to do and things that we hear across customers. But, you know, Shuba, again, kinda looking at those specific, metrics. We've already talked about the… That 20 to 40 person war room that used to exist. But looking at the kind of the… By the numbers, what has EPX Advanced Plus, done to ticket volumes, you know, MTTR? And and, also, you know, what are some of the upcoming goals that that you and the team have for continued, deployment? Right. So, I mean, any any any KPI if we want to measure, right, we would… For this case, we would at least, like, divide it be… By the success rate rate, like, the detection success rate, like, how many times we have been able to detect issues, proactively. How… And the next important part is how many times we have been able to route that issue properly to the right team, which Rohit was explaining earlier. Like, everything started with the network team. So now the approach is different. Now whichever team has the issue, it be… Because of the detection and analysis, we can route it faster to the right team and then ultimately resolution. So, as as far as number of tickets are concerned, it has gone down exponentially as far as the network tickets are concerned. Right? And, we'd lit… I don't think we have the war rooms anymore. So those are the things that, I would say an indicator of how this tool is working. Yeah. And, we have, I I give an example of the collaboration tool. I give an example of a of a executive's house, Wi Fi. In fact, our… There was also our office primary ISP Which had issues, which z d x was able to detect. And the moment we switched over or made the then secondary ISP or primary The issue was gone. So even… Like like, there have been, like, many… I would say, like, it has been, like, few months that we have been using ZDX, and we have act got, at least, I would say, 10 new cases Of of real detection and and and major ones, not like one or two users having issues, but, like, like, broader ones like a office ISP issue and then Perfect. A a application, of another office in, I think, Singapore we detected. So… Yeah. So it has been… I think those successful detections every time has proved the value. And and Absolutely. Right? Like, proactively. Yes. Yeah. That's… That also kind of snowballs into people predicting us, taking actions, and reducing the tickets that potentially could have been reported. Yeah. Yes. Yeah. It's a great great summary. And in terms of what's next, I know we've talked about some of the things that you're looking at, expanding into real user monitoring coverage, you know, and and proactive, endpoint health workflows. You talk about kind of instead of, you know, setting up alerts for those, but then even using network intelligence to, you know, take and detect those ISP outages, but then using ZIA to perhaps, you know, route around them, proactively. Right? Yes. And, I think one of our next upcoming thing to add to this, we would be onboarding. We're working on it to onboard the MCP For a Zscaler, which has the ZDX component. So that would open a lot of automation avenues that we are, as I said, we are… It's it's great to consume the data of the reports because it's pretty nice and easy to follow. But if an agent could be doing that and creating tickets on our behalf with, enriching the tickets with also data Yeah. That that… That's, that's great. Then then another agent for the, service owner team can pick that up in… From that ticket and go towards the final end to end resolution. So that's kind of a broader next goal that we have. I absolutely love that. It continues that touchless from detection all the way through accurate assignment without any intermediary, you know, intermediary, in between. Just an amazing thing, and and we see organizations doing that as as well. Shifting back to you, Rohit, kind of summarizing out. If you had to give some some advice, what would you give to a a network leader peer experiencing some of these same challenges that you've been able to resolve? Advice, I think the the rule of thumb that that network leaders or in general, infra leaders should have is, like, focusing on the the operating outcomes. Right? Not not not what the feature lists for for products are. And so starting with, you know, just a data driven approach, what are the recurring problem categories in in individual orgs? And and ops issues are are distributed. Right? So you could have, like, slow SaaS that an application owner, owner might be facing. For infra point of view, we have, like, home office Wi Fi, cloud path, ISP path issues. Endpoint team have their own app crashes, etcetera. And then, you know, as these tickets are bouncing between teams, I think some of these leaders have a very good insight into identifying those problem statements and then, you know, baselining the ticket volumes, having those categories meaningfully diagnosed. Right? So that, you know, which teams can be pulled in, you know, what's the troubleshooting cost that is getting, you know, accounted for for for the broader company. And so having that that awareness and analysis is critical, you know, before jumping into solutioning. But once you have that data, I think really, you you run a, you know, focused proof of value. That… That's what we did. We picked select number of tier one apps. We call them kind of crown jewels that we wanna, you know, secure, make sure the performance and, you know, telemetry is in place. And then for those, you know, smaller section of infra, we identified, like, what the the the problematic locations might be, right, in terms of globally where our users complaining. And at the end of the day, really partnering with Shuba, we we decided, like, there are three key things we wanna prove. Can this platform isolate, you know, the affected layer, at which the issue is happening? Can we route that issue to the correct team seamlessly so that the owner is identified and acknowledging that issue? And then at the end of the day, is it allowing the team to act, proactively maybe, you know? But even if not, how soon can we act, if this becomes like a major issue that's impacting, like, an entire headquarters site with thousands of people, or an entire region, things like that. Right? So identifying the problem statements, identifying the the cost, and then eventually translating that into a very focused proof of value, That would achieve, you know, finally, what sort of funding model, you know, leaders would have. Because it's not just about the license. It's about defining ownership. We we discussed how other teams are involved. Sure. You need call to action from them to define, you know, what are the the the the endpoint thresholds? What are the SaaS application thresholds? Do they have runbooks and escalation path? If not, we gotta build those ground up as part of this initial, you know, workflow. Data retention was a key. ZDX has extremely crucial data, that is priceless. And so how can we, like, retain some of this… The the… This data, some of these signals, that helps us, you know, just just maintain, go back, identify patterns? So, yeah, overall, I would say my advice is, like, z d x will pay for itself, when network help desk, client engineering, SRE, app owners, agree to kind of act on those signals. Right? If those signals are not, you know, taken into a broader operational workflow, then it's just more noise that… That's that's part of your monitoring toolkit. So without strong cross functional alignment and, you know, clear ownership, leaders might struggle. But if if they follow this model, I think it really worked well for us, you know, part of the… Fantastic. I couldn't have said it better myself. The… You know, a… Understand your current state. What does that set you need? Finding the tool that that can do that in your case and and then many on on on the call, you know, already using Zscaler… ZDX is that most, you know, natural choice given it's already running through that single client connector. It's able to go through, zero trust exchange that other tools can't. And and really that collaboration that is the first step in breaking down those silos and and the kind of infighting between the teams, and now you can really build that that collaborative relationship. Well, I can't thank you both enough for this conversation. Kind of in one takeaway, Rohit, if you had to summarize this this impact that you've been able to have with this journey to Advanced Plus in just one sentence, and I know that's not always easy, but in just one sentence, what what would you say that is? ZDx Advanced Plus helped change the operational game for Docusign. It it transformed our IT support from reactive to proactive. Right? We have proper troubleshooting, routing of issues, and more importantly, I would say, you know, we're automating our way out, to achieve that zero touch operations. So, finally, we we can see, you know, some light at the end of this tunnel. We've been talking about zero touch operations for long, but this tool is is, enabled that vision for sure. Wonderful. Thank you both again. Rohit, Shuba, can't thank you enough. Greatly appreciate it. Thank you. I'm now joined by Luis Rico, a principal product specialist for zDx, here to give us a a demo showcasing what Rohit and Shuba just went through, talking about how they were able to take and proactively detect, all of these disruptions and proactively resolve them before users were putting tickets in and especially get rid of those those pesky war rooms that that ate up a lot of time, from organizational leadership, down into the end users. So in this, you know, five ways to get to root cause faster is the the kind of the pillar of our demo. We're gonna go through network intelligence, managed monitoring, UCaaS call quality, device health events, and even remediation that an l one analyst can do, and the showstopper for for any demo that we give these days, zDxAgent, which as it's been described by some of our customers, it turns an l one analyst into a network telemetry guru. So with that, Luis, I will stop sharing and pass it over to you. Thank you, John. So we're gonna start off here. I believe you can see my screen now. Is that coming through, John? It absolutely is. Yes. Thank you. Alright. So, the new digital experience or the experience center is going to drop you into the menus here where if you have the Zscaler platform, you will see all these other products if you're subscribed to them. For today's, demo, we're gonna focus obviously on digital experience. And as soon as I click on that, I'm brought to the overview summary page. Alright. You can switch the time frame. You can go back two weeks in time, visualizing two days at a time. But as you can see, in the last forty eight hours, what's really nice about this single pane of glass view is that I don't have anything really to be concerned about. No incidents, which is usually when a lot of users are impacted via on the… In the same location, and none of these types of incidents are lit up. I don't have any network anomalies to be concerned with. The applications that ZDX is monitoring on our behalf are good, and everything else seems to be copacetic. So this is a really quick way to go ahead and get a good idea what's going on with the environment as a whole. Absolutely. On demand whenever. And I think the biggest call out there is for standard customers, twenty four or two out… Two day window for forty eight hours, but with advanced plus like we've been chatting about during this webinar, a fourteen day window to look back. But let's dive into one thing specifically. And right now we're gonna start with, a a feature capability that ends all of the I, ISP war room discussions and that is is network intelligence. And one of the challenges that we know that organizations face is that, networks, you know, are, you know, users are are going across networks that you don't control, whether that's because of travel, because of a home, home network. And when Comcast or AT and T has a a brownout in Dallas, how do you find out? Well, usually that's when users start calling in, tickets start coming in. But that's hours after they've been impacted. In some cases, the the issue has been resolved. They've lost a ton of productivity. But what we're going to show you here is how you can proactively detect those issues even before users' tickets start flooding in. Absolutely. And as you can see here, we've got a… Quite a bit of information, but the important part about network intelligence is that, in the past, you would have… Or perhaps with, in some organizations, you have a lot of other different tools to try to get this information, which zDx can already provide. So think about the zDx client running on client connector, nothing new to deploy, and you're getting this additional network visibility as the packets leave the user's environment, right, which was usually the difficulty. Where are the packets going once it leaves that last mile ISP, intermediate ISPs, zero trust exchange for for Zscaler? Where are they going? Without you having to do anything else, we're tracking that for you. We can put it here. No extra tools needed. And as you can see here, we've narrowed it down, for you. The red bars indicate where there were severe anomalies, meaning that there was at least a three times increase in latency or drop packets. Right? And medium is up to a two x increase. So we've had a couple of these severe ones. We're breaking them out by where we have seen users over here on the top right hand side. But when I click on this area, which is of interest to me, on November 5, we had some, several anomalies located in the continent of Asia. So once I click into that, it further drills down and tells me, hey. It's in Singapore. I click into that, and now it drills down and shows me the intelligence of what's happening, you know, with these severe anomalies. Is everybody affected? Is it just… Is it before it hits the zero trust exchange? Is it after the zero trust exchange? And as we can see here, notice this graph down here. Everything is looking fine. There's no red. But if I click on zero trust exchange to application, woah. There we go. We got a couple of red bars showing up in… Indicating severe anomalies. I click on it, expand it a little bit for that time frame, and now you will see the graph below changing to show me the intermediate ISP that is introducing inadvertently or otherwise latency or drop packets for the folks traversing that particular link. And as you can see here, it's Telstra Global, and you can see that the typical p 50 baseline was less than a millisecond, now has jumped up to over seventy nine milliseconds. And if I wanted to know which users are affected, you can see that there's seven users affected. Click here real quick. It'll show me the names of those users, which we could then go into. And then I will show you the IP address view. There's those users. It is clickable. It's actionable. Right? So you can go in there. One thing to note is that even though there was severe anomaly, the severe anomaly doesn't necessarily translate into poor performance for those users. It just means that during the time frame, there was an increase of latency or drop packets. Now if I do IP view, we can show you IP addresses from here to here. This is the link that's affecting these seven folks. I can even say, hey. I don't get that IP, and I don't know where that is. Show me where that is. Okay. Fine. Easy enough. We click on that button. Now you know geographically where they are at. So during this time frame, seven users might have been affected. It was affecting Microsoft login and cloud share applications. But other than that, everything else was looking copacetic. So not necessarily a fire, but interesting to have the data without you having to deploy any further tools using the data that CDX is already producing from your endpoints. Yeah. Absolutely, Luis. And and, you know, one of my favorite stories here is a a global IT services company. They had previously been dealing with twenty four hours of detecting these problematic ISPs mainly by waiting for enough tickets to flood in, to realize that they had a systemic issue. And using network intelligence, they were able to cut that down to five minutes and 98.2% faster, which ended up saving them over $400,000 in in productivity from just one incident by rerouting a thousand users, to, a different data center around that that bad ISP. And I think this is the best time. There's a a question that came in, about that routing around. So, you know, it's one thing to get this proactive alert to say that, the the ISP has an issue, but what can I do about it? So Luis, can you talk a little bit about, you know, kind of just briefly given where we are with time, but, briefly around the value that ZIA can then route around? Yeah. If this had been an actual, like, in the case you described, what happened was they went into ZIA and went ahead and they changed the affected users to use a different sub cloud and took a different path until that ISP fix their situation. Right? So you have at least a way of determining and knowing when things like this happen, take appropriate action. But better yet, you can confidently say to your management, hey. This is not affecting everyone in Singapore. This is only affecting seven users in Singapore, which is a big distinction when people are panicking and you've got a couple of angry phone calls. Yeah. Absolutely. And it's a great way to have the value of what zDx plus, our core core platforms at ZCLR, ZIA, and ZPA can do together. Thanks so much for that, Luis. Now that talked about handling ISPs after users are online, but what happens before they start their day? One of the things that with with zedX standard, you're able to monitor apps, by the users who are actually using them. But what about monitoring, you know, Salesforce Force before the Tokyo team comes online and starts their day? Or for example, the the service desk team using ServiceNow in Frankfurt before their EU shift starts. And one of the things that we have the ability to do, with managed monitoring is to synthetically probe applications twenty four seven so that ready when they are to come online. So, Luis, take it away. Absolutely. So, for this example, we have a very, very good service for monitoring those services, APIs, or websites or web applications that are critical to your organization or maybe to its to your partners, without having to put that burden on your employees' endpoints. What do I mean by that? You want your employees to use the application to be able to monitor their applications that are, important for them for production purposes. Right? But you don't need them to be monitoring your website constantly to determine that your website that serves the public is okay or that a service like Genesys Cloud is available or not available. You don't wanna necessarily put it on the endpoints. Now you can, of course, but what this gives you is, for example, Genesys is a voice over IP platform that is very, necessary for a lot of organizations. And what we do here is we tell our Zscaler data centers. Right? So these are the data centers that are available, to to probe from… To go ahead and probe the Genesys cloud, do the page fetch time or the get request to that, as well as do the trace routing from the individual data centers that we have and even give you a visualization of the times and the latencies that are observed going to those services. What's the value in this? Again, not having to wait till everybody wakes up and turns on their desktop or their laptops in order to start getting some telemetry, you… We are doing it on your behalf, and we can show you that in five minute increments much like the other data in ZDX already. So managed monitoring is a critical part of monitoring those services outside of your organization without requiring your employees to give up, performance on their machines just to monitor these outside services. Yeah. All all great points. The one thing I'll add is that these, synthetic probes are coming locally. We have, data centers around the globe so that we're able to, simulate your actual users. It's not just always coming from Melbourne, Australia or, you know, Frankfurt, Germany. It comes, in the areas that you'd expect them to come from. I was on a customer call last week where the light bulb came off here of, oh, you mean that I can get an alert to my team at 03:00 in the morning that an app is down and have a fix in place before the the manufacturing, you know, business opens up at 6AM. You know, how how cool is that from from a use case there? So we're on a a video call right now. It's the way that modern businesses operate in this hybrid international world that we are living in. So let's talk about the thing that seems to never go away, and that is poor call quality in applications. And this is one of our favorite things to talk about because especially when someone's working from a hotel or a, you know, a home WiFi, it's very easy to blame that network that we don't control, but it doesn't solve the problem that the users are facing. And more importantly, it doesn't allow the business to operate, especially when you have executives that are globetrotting around the globe. That's a little redundant, but globetrotting around in their travels needing to join meetings with, board members or, you know, c suite, they need to have good call quality. So let's talk about that next, Luis. Absolutely. So CDX Advanced Plus gives you the ability to tie in the call quality, APIs from Webex, Zoom, or Teams. Right? And in our example, we're gonna use Zoom as an example. What this allows you is I like to think that it allows you to offload the responsibility for those level two, level three calls regarding problems with Zoom to perhaps being able to empower your level one so that they can start handling some of this stuff, at least triaging it and getting it to the proper resource instead of just simply throwing it over the fence and letting the network team handle it. Right? So this is a big win in terms of ZDX advanced plus and being able to give people that ability to help quicker. So I've gone over to Isaiah, and Isaiah was the host on this meeting having poor score. And you can see Zoom call quality is different than the other applications. I'm just gonna select them and show you real quick here in the visual so you can see, but the data is pretty much set up the same way. So these are the other applications that he has that he's monitoring from his system. They all exhibit pretty poor scores, but we are focusing on call quality here. And these, this dot here represents the beginning of the call and the end of the call. Right? So that's what these little dots represent as call quality. Now as you can see here, we see that his last score, four. We are expecting a four or higher. The meeting metrics from here on down are coming direct from Zoom. So this is being retrieved every five minutes from the Zoom call. Right? So we have this, and the latency, horrible. Four hundred and seventy eight milliseconds. Anybody who knows anything about call quality or voice over IP, above one fifty, you're not gonna have a good good experience. Right? So this is pretty bad. Forget about video. There's no way he's gonna be able to do video. Sharing quality, terrible as well. Right? So what do we bring to the table? Well, we got those statistics from the… For Isaiah from Zoom directly. What we do is we marry it to the data that ZDX is constantly probing for. Like much of the other applications is we're giving you the path from it and much, notice that there is no z I in the mix, which is exactly what all these vendors recommend. No VPNing, no hairpinning. Don't send through a security service. Send that traffic direct. It's voice and video. There's nothing to secure there. Just send the direction of best experience. We can prove it. It's right there. However, there's a big problem. There's your latency. The user's on prem. It's an Ethernet connection. I can now, as a level one, say, hey, mister Isaiah, I see where the issue is. I have to escalate this to my networking because it happens to be in between the gateway and the egress, and there's nothing you can do in that, unfortunately, unless maybe you wanna switch to Wi Fi. But I will escalate this to networking, and they go ahead and escalate to networking and move it along. Imagine being able to solve problems that quickly as opposed to just saying, oh, it's a problem with your call quality. Let me go ahead and send you to a specialized team who then has to pull I don't know how many different tools to get this information that we can get in a couple of seconds. Yeah. All all great points. And it kind of gets to the the question that we we had come in during this this segment of our network team says call quality issues are always the ISP and our ISP always says the issues are us. Absolutely. How does z x… How does z d x settle that? And, one of the things that I always love to, you know, hear and and and Shuba mentioned is as important as knowing what is the problem, even knowing what isn't the problem is more important because it means the ticket is getting to the right team the first time and it's not bouncing around while issues still keep on happening. So, we've seen the ISPs, we've looked at monitoring applications proactively and now we've even driven into to call quality. But what about the 30,000 devices that are on your network and they're slow and get ting slower. And what we know about organizations is, you know, right now the way you're typically monitoring and managing these is you've got the service date and you're kind of guessing of what the device health is. You're, you know, reactively responding to tickets. The service desk is getting on, you know, trying to, you know, figure out, okay, CPU spiking. Why is that the case? So you got all these blue screens of death. What is going on? And with our device scoring and, device overview health dashboard, we have the ability to give you an entire fleet wide view of all that, and to proactively manage it at the same time. So, Luis, take it away. Absolutely. Now there's a lot in this, overview devices, Luis. We won't have time to go over every single, tab or feature. However, like the… It says here, this is a great overview to understand how is… How are the endpoint devices in my organization doing as it relates to hardware performance. This is not being measured against any application in particular. This is hardware related only. As you can see, 98 out of of the eligible devices during this time frame are… I have about 20 of them, almost 20, that are listed as poor. Why are they listed as poor? You click on here, and now we list those, and we can show you, hey. This one here has been listed as poor due to poor CPU utilization. Memory usage is probably through the roof and on and on with these devices. I can, of course, click on these things and pivot over, and show you the individual information. But at the moment, we're gonna go right back to here and continue with the other bits of information that are very helpful from a fleet wide perspective. As you said, it's important for me to determine also, hey. Is my refresh cycle making sense? Do we have the appropriate endpoint devices with the usage profiles we have? And we do a pretty simple calculation that helps you determine, hey. We got about 10 power users who are still on low power devices. We should probably go ahead and, upgrade those folks first. If you need to find out which users those are, you can click here, and we'll be able to show you those users, export them, maybe help those folks out since they're probably, in bad need of getting on some new hardware. Now there's also these other tabs over here. We won't have time to get into them, but we are collecting a lot of very useful information regarding the devices that will help corral a lot of the things that you may not have known before regarding, your devices as a whole. Yeah. Yeah. Everything from crashes, blue screens of death, the applications and iOS, you know, additions that they're running, all of that information to be able to to look at full visibility across. And one of my, you know, favorite stories here, you know, I happen to in product marketing, I typically am in one of those under provision because I've got, at a minimum, a 100 Chrome windows open at any any one time. So my machine is always struggling. But in in one particular instance, a large health care system, you know, we can't give their name because of privacy requirements, but they were, kind enough to let us share the story. And through this, device scoring capability, they were able to identify 14,000 blue screens of death, 10,000 audio failures, audio software failures, and fifty thousand browser crashes. That's five zero with three after it over just thirty days. And what that meant for them was the ability to avoid, what they measured as $606,000 worth of productivity across 15,000 employees. So a massive, massive, impact. And as we started this talking about, Gartner gives that stat of $336,000 an hour to the average organization in lost productivity. Now, all the things that we've talked about so far have given you the ability to proactively detect disruptions, get alerts, and it's one of the nice things with advanced plus you can set up a 100, proactive alerts by geographies, by user types, you you know, applications, network, device, all these different things that make sense for your organization. But what happens when a ticket comes in? That's… Has to first get triaged by the l one analyst. And, you know, some of those telemetry points that that Luis went through, you you know, he does this day in and day out. He is, one of our gurus here, on the ZDX staff, and he knows what he's looking at. He can translate all of those and understand actually what's going on. That l one analyst, most of the time it's guessing. I know we have a lot of network people on the call, and typically you become that easy scapegoat. So, Luis, let's unveil, ZDX agent and show what this allows through simple natural language that l one analyst to go from vague ticket to root cause in the matter of a few minutes. Lulu. So up here in the upper right, you see the magic wand that says eAgent. When I click on it, it gives me this little side window. I'm gonna go full screen. I had already ran something here before because I wanna keep it, short for the time frame that we have, but it's a very simple query. So imagine I'm on l one. I've never touched z d x. They're not me. They have no idea what to do with all this stuff, and they just got thrown a difficult case as, hey. Listen. I'm sorry, but you gotta go help this person. What do I do? Well, all I did was I typed the user's name, troubleshoot Barry Bogdan for Zoom issues in the last twelve hours and let it rip. That's all I had to do, and it comes back and runs this complete analysis and tells me, hey. We had problems with the score. Yes. The CES score for the applications went down. We also saw that it had issues with device health. Right? Adobe Premiere Pro consumed the highest ninety fifth percentile. Right? So Zscaler tunnel and Adobe Premiere Pro, Adobe Premiere Pro are 47%. So it's basically calling out Premiere Pro, and it happens that when Premiere Pro is running, the ZDX score drops. As you can see here, it's… The ZDX score drops and gives them poor performance. Now it gives me, sorry, gives me additional information that also there were times when the network was also… So it's two things for Barry, not just the CPU because you can have high CPU utilization and not necessarily have high latency on the network. But you can have both I… Both things occurring at the same time, making things double worse. So we have two things to look at. The level one can handle the the issue with Adobe Premiere Pro because he is a creative and say, hey. Look. It seems that at times, it it runs a little hot. Maybe close some of the jobs you're doing in Adobe Premiere Pro. As for the latency, you know, I'm gonna have to, go ahead and escalate that to the network team because we're not… It it is on our network, on the on prem network, and I don't have purview to help you with that, but they can probably go ahead and help you with that. So they don't have to be a superstar or rock star. They can just ask it and not just simple questions. You can… We can go in deeper. And for a lack of time, we're not gonna go in deeper, but we can ask really pointed prompts about what you wanna get from the user data that's available in your organization, and it does a hell of a job of giving you an analysis on what's going on. Now to keep it, the last little bit is should you be so curious as to see what buried bogged down looks like? We also give you the option to, hey. Just click right on in there, and you're gonna be taken to the very familiar user details overview page, which has been a staple of ZDX since day one. So we haven't taken anything away from you. It's still there. It just allowed me to go ahead and do a real quick pivot, go to that user after the z agent at the end, and now I can actually look at the data for myself as well. So Yeah. That l one analyst took and filtered through all the noise of symptoms that Barry thought were the the problem, and was able to get to that root cause. So now we can can remediate that right there. Luis, thank you so very much. It's been a action packed twenty minutes or so here. We've gone through ISP, through through network monitoring. We've gone through application. We've gone through UCaaS quality. All these different things. I'm going to take over the the share here, and finish us out here. And the the last thing, you know, wanted to kind of wrap everything up and and showcase what we've been able to to do here. So, when we look at all of the the customer case studies that we have across ZDX, they have a few things in common. They have a 52% reduction in the meantime, to resolution, 330% ROI over three year time period through loss, product… You know, gains in productivity, and the average enterprise organization because of the gains in productivity, the consolidation of tools, $1,700,000 annual enterprise savings. Just an incredible value. Before we, wrap up, wanted to thank Luis for his expertise. Wanted to thank Rohit and and Shuba from from the Docusign team. But don't just stop here. Cut… Take the next steps. You've taken the time to watch this this webinar. There are a few different areas. We have some great ebooks that you can, you know, download. Go to our website and see the customer's, case studies in more detail. But one of the things that I think is most valuable to click around this yourself, sign up for one of our upcoming workshops that we have so that you can actually play around with these, tools, see simply how easy they are to use and how they have impacted the organizations that we've talked about. Again, thank you to all of our presenters. I hope to see you on our next ZDX webinar. Thanks and have a great day. Scott Bekker Alright. And, Arena's opening up a. short Video you for joining Scott Bekker we'd really appreciate if you could all fill out. Video our our Scott Bekker for, those of us on the actual tech media team. Video solves network app Scott Bekker you're filling that out, I'll just mention Video stopped Scott Bekker couple of things there in the doc. section. Video Let's jump right into our agenda. Scott Bekker that z d x practitioner's, guide. There's Video cover Scott Bekker virtual Video right. now. Feel Scott Bekker the interactive Video ask, questions Scott Bekker there's the Siemens Video way. Scott Bekker customer story. Video We will Scott Bekker be sure to, grab those Video we are able Scott Bekker the, Video or we'll, Scott Bekker the waning moments Video them, Scott Bekker of this event, Video and follow-up, afterward. Scott Bekker before those links disappear. Video We'll go through some. introductions. Scott Bekker And, Video We'll go through Scott Bekker least, Video what happens Scott Bekker do our prize giveaway. Video evolve Scott Bekker winner of. a $250 Amazon Video have a great conversation, Scott Bekker Adam Fouch Video two guest speakers. here from Docusign, Scott Bekker congratulations to Adam Fouch. Video their story and, Scott Bekker always, we'll be in touch about, claiming your prize Video demo Scott Bekker the event. Video a twenty minute demo, Scott Bekker with that, Video five ways Scott Bekker of the actual tech media team, Video to root cause Scott Bekker all of. you for sticking with us today. Video let's Scott Bekker was great having, you all here for this actual tech. media event. Video Rohit Scott Bekker look forward to seeing, you again next time. Video engineering manager Scott Bekker have a great Video infrastructure. Scott Bekker rest of your day. Video Rohit will join us here along with, Shuba, Chakrabarti, who is a lead site reliability

Please provide citations & attributions for any information you generate from this content, thank you.