Matt Roberts, healthcare lead, Juniper Networks

Unleashing the Potential of AI-Native Networking

Industry Voices AI & MLNetwork Automation
Matt Roberts

HIMSS Healthcare Webinar: Unleashing the Potential of AI-Native Networking

In this webinar, Matt Roberts, Juniper’s healthcare lead, and Sunalini Sankhavaram, VP of products, share how healthcare ecosystems are evolving to address the complex challenges of care delivery and integration of complex systems. With AI and AIOps, transformative capabilities unveil operational efficiencies, elevate experiences, and foster sustainability, ultimately shaping the future of healthcare.

Show more

You’ll learn

  • How seamless connectivity boosts operational efficiency through reliable interoperability, resource optimization, and secure exchanges

  • The importance of modernizing and optimizing the network to improve integration

  • Simplifying complex healthcare operations with AIOps

Who is this for?

Network Professionals Business Leaders

Host

Matt Roberts
Matt Roberts
healthcare lead, Juniper Networks

Guest speakers

Sunalini Sankhavaram Headshot
Sunalini Sankhavaram
VP of Products, Juniper Networks

Transcript

0:00 hello my name is Patty and R from hyns and welcome to today's webinar thank you for taking the time to be here today we

0:06 have an exciting program for you but before we get started I would like to highlight a few things so you know how

0:11 to participate in today's event at the bottom of your screen are multiple application engagement tools you can use

0:18 all the engagement tools are resizable and movable so feel free to move them around to get the most out of your

0:24 desktop space if you need technical support during today's event click the question mark icon and the navigation

0:30 menu at the bottom of your screen to find answers to frequently asked questions or you can type your questions

0:35 in the Q&A widget to the right of the media player you can use the same box to submit questions to the speakers at any

0:42 time and if we don't get to all of your questions today we'll follow up with you offline engage with today's content and

0:48 let us know how you're feeling with the Emoji reactions if you miss anything don't worry you'll receive an email

0:53 within 24 hours with a link to the on demand recording when it becomes available so now let's get started with

1:00 today's program unleashing the potential of AI native networking sponsored by Juniper Networks our speakers today are

1:08 Matt Roberts Healthcare lead andolini saram vice president of products at

1:13 Juniper Networks with that I'll hand it over to Matt andolini thanks Patty

1:19 welcome everybody today sunin and I are going to touch on a few Trends and challenges within what I'll calls the hls segment and how AI Ops is elevating

1:27 user experience and uh also overall it efficiencies in these particular IND

1:32 Industries so uh let's jump right on in here so as many of you likely realize

1:38 healthcare industry faces a large amount of challenges across the board today I'm probably uh putting up a few statistics

1:45 that you're all very very well familiar with at this point in time uh however you know not only are we concerned with

1:51 the bottom line here in healthcare more than ever there's so many moving parts to take into account as you look at

1:56 these particular segments cyber attacks continue to increase we've had over 133

2:02 million patient record breaches in uh 20 uh 23 which is an absolute record I

2:07 think it was 156% increase over 2022 nursing burnout unfortunately continues

2:13 uh you know it's significantly been uh ticking up if you will since covid due to a wide range of different factors

2:19 could be expanded duties uh increased patient workload uh also safety unfortunately comes into play here too

2:26 uh let alone the time and energy having to search for equipment uh which is not only annoying but it does have impact on

2:33 patient care that also has impact on morale within the facilities and you add in layer zero things and energy

2:39 challenges that continue to mount as the cost of power and cool buildings infrastructure all those things continue

2:44 to uh to take place and we have to be cognizant of that now if we add in the emergence of AI and I fully realize AI

2:52 it is a buzzword right with everyone having their own definition of what AI truly is but over 3ars of healthc Care

2:58 Professionals expecting widespread adoption over the next three years it's becoming very very real okay and so

3:04 there's some really yeah what I'll say is tangible benefits that I hope you get out of this today uh that can be

3:10 realized uh but we also have to be cautious when we talk about privacy we talk about governance we talk about

3:15 ethics and so if sunen and I uh do our jobs right today which I hope we do uh

3:20 we hope you come away with actually some several good use cases where AI is fundamentally making a difference in

3:26 these types of environments so a few more trends that we felt were interesting especially as it relates to

3:32 outcomes experience cost reductions they're all much needed in healthcare you'll see those I'm not going to read

3:37 each and every one of these uh you know on the slide here you'll see AI mentioned multiple times and I firmly

3:43 believe that you're going to continue to see this technology become more entrenched in all aspects of digital care okay and so if I'm being honest

3:50 it's also all about the data all systems obey the Golden Rule I think we remember

3:56 this uh if you took Computing classes garbage in G out so if you want good results you've got to feeded good data

4:03 and this is why the quality of what we like to say is the Telemetry data really really does matter and as we go through

4:09 the presentation today my hope is that you'll get a good understanding of the seven plus years that we've spent with

4:15 data science and training models that are in place to really produce solid results for our customers literally

4:21 changing the day-to-day operations for the IT staff now not only are each and

4:28 each of these three pillars uh they're fine it's a it's a fine delicate balance uh it's an absolute necessity in today's

4:34 Healthcare world uh I think we all understand the number of devices the number of users digital Solutions

4:40 they're not going down right it's only going up I think that's obvious probably to everyone listening in the plan is to digitize everything from high dep video

4:47 to robotic surgeries to AR VR uh 640 slice scanners you name it right the

4:54 technology just continues to be more advanced so it's either here today or

4:59 it's coming in your environment today so you have to be prepared for it right you've got to prepare accordingly and to do so it's a balancing act uh keeping

5:07 the lights on while planning for strategic growth right so that kind of fits right there in the middle now

5:12 migration without disruptions it's a necessity in healthcare uh we know this is a 247 365 environment um it is the

5:21 difference between also allowing patients to be admitted seamlessly and cleanly versus diverting patients to a

5:29 competitor because your systems are down we're fortunately we're seeing a lot of this stuff today and all this is

5:34 underpinned by the ability to safeguard the data uh when we start to talk about against cyber criminal activity okay and

5:42 so you've got to be able to do that at all costs and as mentioned on the last slide there cyber criminals are getting

5:48 smarter for organizations it's sadly no longer if uh but it's when a compromise

5:55 takes place and that's why it's also important to be able to detect and respond at a much much higher speed and

6:01 if you can build predictability into the infrastructure to do just that you're going to be in a much much better position right when those events

6:07 unfortunately do take place so while challenges continue to mount there is

6:12 light at the end of the tunnel it's critically important to also not lose sight of how everything is connected in this digital era right so this is why

6:18 we're here today and we're talking about this and it's been proven time and time again that connectivity is not the same

6:24 as experiencing a great connection and I'm sure most of you that are listening in today have never heard any of these

6:30 complaints that are put up on the screen you know we've got a great fast you know connection but the EHR is still slow why

6:36 what's wrong with my tele Health session why is it jittery why am I dropping packets or I'm getting half of a voice call working or I'm dropping as I move

6:43 around the facility from room to room or from hallway to hallway right A lot of these things that we hear and joking

6:49 aside we we we all hear them and we continue to hear them but let me ask a simple question to everybody if you

6:55 could take a big big Spotlight and shine it on the dark spots of your netor work would you try

7:01 it okay so as we go through this today I want you to think about that and I mentioned a spotlight for a good reason

7:07 so if you're you're in healthcare it operations you fully know that the environment is filled with blind spots

7:13 and it's a lot of this finger pointing between teams right and if the connection experience isn't measured you

7:19 absolutely have no idea what type of user experience is actually being had so

7:25 how do I know if my clinicians my staff my patients my guests they're having a good experience how do I know if my

7:31 applications are performing properly which may impact in user experience why do I rely on my network staff or my

7:39 users to report the issues back so that my network staff then has to jump in or my help desk to try to troubleshoot

7:45 something which can take not only minutes hours can take days and weeks to try to figure out the issue sometimes

7:51 things don't get figured out and we call that chasing ghost and you're continually chasing a ghost and then

7:56 also you know how do I reproduce issues much much faster because today it's like

8:02 digging sometimes in a needle in a hay stack and trying to find that needle and so again as we go through this presentation today you're going to find

8:08 different ways to leverage AI Ops to do just that to fix these these issues that are very very common and we'll call our

8:14 blind spots within the network um the good news is is that we're not asking you to be the first you know provider

8:20 life science company on the Block to try it and so what we do with AI um I'm going to keep it in really really simple

8:27 terms is we help our customers deploy netwk Works faster lower costs fewer

8:32 mistakes right those three these three pillars uh while we also provide day two operation advantages so

8:38 lowering uh and seeing 80 to 90% reduction in trouble tickets so again

8:44 how do you know if your clinicians are having a good experience on their on their devices uh on their applications

8:50 like z uh tomb excuse me if I could talk teams or Zoom barcode scanners well if

8:56 we can take the guesswork out of network experience and we can use reinforcement learning and AI technique all combined

9:03 with realtime feedback it's simply not possible in a traditional controller

9:08 based architecture so you've got to do something different through Modern microservices which soon ly will address here in a few minutes now we get

9:16 Telemetry data from teams and zoom so if your physician had a a zoom call it was

9:22 a bad Zoom call last Thursday and it was perceived to be the network let's say R AI would have told you just that okay

9:29 however with video conferencing it's not always the network even though it is the first thing that is oftentimes to blame

9:35 so now that we've got the Telemetry data from the endpoint on teams or zoom and those particular examples uh we're

9:41 finding out frequently something actually maybe on the user device that's pegging the CPU which was maybe impacting the user experience so again

9:48 getting down deeper into what the application is what's happening on the network and then triangulating all this

9:54 data and getting to the point and resolving in a much much faster time so that the end user can be uh much much

10:01 more advanced if you will and seamless as as we move along now how do we do all this soonly we'll talk a little bit more

10:07 in detail but we capture over 150 different state changes and data points on every client every minute uh we know

10:14 if somebody had a PO roome uh we know po rooms are very common in healthcare we

10:19 know if they you know if they or someone typed in a password wrong so we can't see the password but we can see that they failed the authentication step and

10:26 we can see why uh we signal to our cloud uh and we ask AI to do what your best

10:32 technician could do if they were sitting there 247 side by side with every clinician Walking The Halls with

10:39 sniffers really trying to watch The Experience okay so we're taking a lot of the manual work out of this additionally

10:46 our virtual Network assistant which we call Marvis continues to be what all also say is a huge cya for our Network

10:53 opy raters you know today so every day every week every month every year Marvis continues to get smarter continues to

11:00 learn continues to help really proactively identify resolve issues before they actually impact user

11:07 experience so with all that being said I'm going to turn this over tosini who's going to give you a little peek behind the curtain to actually how we make this

11:15 all happen thanks so much Matt really appreciate that overview so again

11:21 welcome all thank you for your time today and let me now tell you how healthc care providers are changing the

11:30 experience for the end users leveraging aops and specifically what we're doing

11:36 here at Juniper as we are transforming the networking industry by simplifying the day to operations not so much from

11:43 your managing your network and managing the boxes but it's all about delivering the best experience across the

11:50 Enterprise stack to deliver the best experience your end users your nurses

11:56 your clinicians your patients can have when they connect to the network and a key part of that um the

12:02 foundations of how we deliver that are all based on the client level visibility

12:07 so as Matt mentioned we create a state machine for every client now why is that

12:12 important why does that matter it matters because that's what allows us to know exactly what the prec connection

12:19 postc connection experience for every user Wireless or wired across the van is

12:26 and if the experience is not good more importantly be able to proactively notify the operators hey Marvis found an

12:35 issue it's impacting these users these users could be nurses these users could be devices that are being used by your

12:41 patients to access applications and not only do we tell you proactively hey here

12:47 are the clients's impacted we also tell you exactly why they're impacted and what do we need to go fix to resolve

12:53 those issues this all goes back into the aid driven support right as Matt said

12:58 you know no longer need to have people go on site when an issue happens to try and figure out what was the issue wait

13:05 for it to happen again set a packet sniffers to try and capture the problem

13:10 with the AI driven support mechanism where we're using AIML constantly in the back

13:18 end it's monitoring all the right Telemetry coming in through our Network

13:24 for the end clients we already will be able to to identify the experience of

13:31 the end users and be able to proactively notify you if there's an issue that's impacting the end user

13:37 experience all of this imagine the zettabytes and zettabytes of of

13:42 telemetry information coming into the cloud for every user connecting to the

13:47 network it's all possible and viable because of the microservices cloud

13:53 architecture that is the underlying Foundation that is essential to be a

13:59 able to provide the client level visibility to be able to measure the

14:04 service level expectations of every user as they connect to the network and

14:11 continue to use the network for accessing their applications and last but not least

14:16 critical to the end user experience also is digital engagement so as part of this

14:22 SAS platform solution and our Network portfolio we also are a able to enable

14:31 realtime engagement use cases for delivering better patient experiences to

14:36 navigate their way around uh the healthcare facilities as well as Drive

14:42 operational benefits by being able to find your assets leveraging again the

14:48 same infrastructure and the same Cloud platform end of the day it's all about

14:54 how we can deliver the best experience to the end users connected to the Network and also make it extremely

15:02 simple for Network operators to deliver that experience leveraging the aops

15:07 network stack that we have built as part of our platform and all of this is 100%

15:13 open API driven architecture it's all standards based again but the focus

15:18 again is all about delivering that best experience the key aspect the key

15:24 paradigm shift that we are driving like why is AI native networking important why do we care we care because we are

15:32 now moving towards managing the end user experiences and not managing the network

15:39 elements as Matt said earlier right making every experience count making

15:46 every connection a great experience where we are moving the needle where we

15:52 are changing the industry is and where we think it's absolutely critical especially in the Life Sciences

15:59 Healthcare segments is how do we ensure how do we assure that the experience

16:06 that every user has on the network is the best possible and that is where the

16:12 whole focus is it's all about client to cloud or client to Applications right as

16:20 your life sciences caregivers your nurses your clinicians access their

16:26 applications access patient records how do we make sure they have the right data available to them at their fingertips

16:33 within seconds how do we make sure that when a patient is is setting up an appointment uh with their caregiver that

16:40 they're able to navigate all the way from their home coming to the parking lot and then most importantly once they're in the parking lot how do they

16:47 navigate huge Health Care campuses to get to their appointments in time and get again the right information at the

16:54 right time VI digital engagement and that is where this a stack comes into

17:00 play removing all the blind spots what is critical here again is the Enterprise

17:05 blueprint what we are delivering with the microservices cloud architecture is

17:11 the ability for us to measure using AIML using unsupervised machine learning

17:17 using reinforcement learning in various aspects of the Telemetry coming in is

17:22 measuring first of all the end user experience as the end user's traffic pattern goes through Wireless the wired

17:29 the St van and we've also now extended it towards our routing stack as well as

17:35 pulling in anomalies from seen in the data center tied to that is the ability to securely on board devices both users

17:43 and iot devices but through this entire network stack how do we assure the end

17:49 user of the best experience possible and that's exactly where continuous learning comes in that is where the Marvis aops

17:57 framework comes in where we're able to measure the experience root cause or

18:02 identify whether it's good or bad and most importantly if it's bad why is it

18:08 bad we are out there every day every minute every hour finding the needle in

18:13 the haast stack problems across the full Network stack that are impacting end user experiences so

18:21 the new kpi of the network for us is not Network up time that's a given the new

18:26 kpi of the network for us is all about that end user client experience are they

18:32 having the best experience possible on the network and if not why is that and how can we then leverage our AIML

18:39 technology to go into a self-driving mode and tell you exactly where the issue might be that's impacting the end

18:46 user now how do we do that how does aops enable such a framework to give

18:53 this granular visibility on every user experience coming into the network and

19:00 that again is exactly where the microservices cloud architecture plays a great role with the

19:08 microsof cloud architecture we essentially have built a platform where we can consume a synchronous Telemetry

19:16 coming from the network for every Wireless and wired client and this is

19:22 where you see us looking at pre- connection metrics on the wireless side such as time to connect successfully

19:28 connects coverage roaming and also very clearly be able to identify hey first of

19:35 all is the experience good or bad so what you're seeing here is a huge Paradigm Shift it's not about just a

19:41 simple time series view of network up time this is about actually measuring

19:46 the network experience in terms of successful user minutes or unsuccessful

19:52 user minutes if we find a user who's not able to connect to the network

19:57 proactively the aops engine will notify you that hey 1 2 3 4 XYZ users are not able to

20:04 connect but also tell you exactly why is it a DHCP problem is it an authorization

20:10 issue is it because of a van circuit issue that we are seeing intermittent connectivity to the clients and most

20:17 importantly once they've connected our service level experience framework is also able to

20:24 identify is every user getting the coverage they need and and even more so

20:30 in today's world it's more about capacity is every user getting the

20:36 capacity they need on this shared spectrum that we call Wireless right today when things don't work the first

20:43 finger that is raised is the Network's not working right and obviously superlative terms in the negative have

20:48 been used to qualify that experience the goal for us is first of all even before

20:54 that first finger is raised on the network being an issue be able to qualify

20:59 is there a good experience or a bad experience happening on the network and then secondly very quickly be able to

21:05 qualify is it a network problem or something else and if it's a network problem exactly where it is so we

21:11 believe in something called meantime to innocence because as many times the network is the culprit it's it's as many

21:18 times also not the culprit um and that's where the meantime to innocence framework comes in with a

21:24 service level expectation so for every user why or Wireless we are creating a

21:30 state machine pre- connection post connection to give you this enormous visibility into what is the experience

21:37 of every user and if it's not a good experience proactively notify you with automatic root cause identification what

21:44 is the reason why and in this journey just like you saw uh you know saw me showing you the full Network stack we

21:51 also leverage Telemetry from third-party systems and that again is because of the fact that we have a microservices cloud

21:57 architecture we can consume Telemetry from zoom and from teams it's a cloud to

22:03 Cloud integration we have built so that we can now very clearly identify if you had a bad Zoom or teams call was it

22:09 because of a client issue was it because of a network issue was it because of a circuit issue on your on your vanite or

22:15 was it because of a blip on the zoom and teams application themselves no more

22:20 blind spots end to endend visibility for a user and the applications that they're

22:26 leveraging for productivity with then aop stack like that we offer and again

22:32 this is across the board this is across Wireless this is across wired you know free connection post connection

22:37 experience minutes for the full stack and on the STD van side as well if you extend our aops platform to the St van

22:44 with our portfolio you're able to see again the connectivity the availability

22:51 and the performance of the user experience whether it's a client issue a

22:57 network issue or something beyond the network like a van link problem or an application Health being the reason why

23:04 a user is not having a good experience so no more blind spots no more finger

23:10 pointing a second guessing with an a op stack that gets the right Telemetry into

23:16 the right Cloud architecture you're able to get full visibility on the end user

23:21 experience but then comes the point of hey it's great that you know an aop stack can give me such visibility

23:29 how can I get notified proactively of the issues that are found in the network

23:35 or outside the network let me now introduce you to Marvis your virtual Network

23:41 assistant Marvis essentially is again driving a lot of day to benefits for our

23:49 customers it is helping simplify the operations of the network so not only do

23:55 we make the Day Z day one deployment of Networks really fast with with our apis and our

24:01 automation but when the rubber hits the road when it's all about you've deployed the network great job but now you just

24:08 start connecting right your portable MRI machines your portable x-ray machines your IV pumps are connecting to the your

24:14 network you're deploying more and more iot devices um you're giving more and

24:21 more client devices to your caregivers to again deliver a better patient

24:27 experience um as end users all of those need connectivity and that's why the

24:33 service Le expectation framework comes in to give you that visibility but when there are issues that is where Marvis

24:39 comes in and proactively will notify you will proactively create a ticket for you your ticketing system because again it's

24:44 all API based to let you know hey there is a bad cable on this particular switch

24:49 on this particular Port that is impacting these Downstream end users there is a missing VLAN often times a

24:56 ticket will come and saying hey I went to this part of the building and my voice call dropped the Network's not working

25:03 there's a roaming problem or a coverage problem often times in fact nine times out of 10 it turns out that there was a

25:10 misconfiguration and the voice VLAN was not configured on that Port where the AP

25:15 was plugged in so when a user roam to that AP the traffic was black hoold it

25:21 was not a coverage issue it was not a capacity issue it was not a roaming problem with either the network or the

25:27 client it simple thing as a configuration Miss but now those can be

25:33 proactively caught and identified automatically with the AIML stack that we have as part of our Marvis

25:39 Engine things like um DHCP things like DNS things like radius when clients are

25:46 not able to connect or they were able to connect initially but you know the DHCP lease was not renewed for some reason

25:53 and now the user who was happy previously is not able to get a good connection all of those can be proactively identified via Marvis

26:01 actions framework to let the network operators know hey you are having an issue on your network that's impacting

26:07 these users and this is the exact reason why and that allows you to again do two things first of all meantime to inance

26:13 you know exactly that it's not a network issue from a wireless wired or van perspective but more importantly it's

26:21 lowering the meantime to resolution so identifying that the users having a bad experience and telling you exactly where

26:27 that KN in the h stack is you can get the issue resolved is again where the key benefits of an aop stack like the

26:34 Juniper uh Miss platform comes into play

26:39 now the other key aspect is how do we see network operators interacting with

26:47 the network right CLI and dashboards are becoming a

26:54 thing of the past just like you interact with your Alexa or hey Google that's exactly the way we see operators

27:01 interacting with the network by a conversation assistant so we had a conv

27:07 assistant that we launched back in 2018 which was able to provide documentation search which is able to provide troubl

27:14 shooting capabilities for your network so troubleshooting an application and you'll see more of this in my demos

27:19 troubleshooting users all that is part of the Marvis conversational assistant

27:24 interface Journey what we've also done recently is put this on steroid steroids by integrating a large language model

27:32 where now we can give you even more comprehensive summarized information on

27:39 complex questions like Hey how do I set up a campus fabric hey I see the aped blinking Amber green what's going on hey

27:46 how do I set up a new network how do I set up a secure connectivity for my iot

27:51 devices coming onto the network and this is where the Marvis conversation interface is able to give you a very

27:57 comprehensive answer and also give you the documentation that is relevant to achieve those use cases leveraging

28:03 llm and it also helps in ticket deflection before you even have to open up a ticket via the Miss portal you can

28:10 now actually ask Marvis right away if you have a question to say hey Marvis help me with this issue I'm seeing and

28:15 Mars will be able to give you an answer as to exactly what you're seeing and also give you an FAQ uh Links of what

28:21 could be additional resources available for you to solve that problem

28:28 where we took Marvis next was leveraging continual user experience learning data

28:36 so I spoke to you about zoom in teams what we built there was essentially the

28:42 ability for our Cloud architecture to ingest user label data from the zoom and

28:48 teams cloud and very quickly be able to identify if you're having a good team call or a good zoom call or a bad team

28:56 call or Zoom call and most importantly if it's a bad team call we have now

29:01 recently launched a predictability and explainability model of why your

29:07 experience is good or bad so essentially we are taking user label data from the

29:12 zoom and teams Cloud we are merging that up with the network tele data we already

29:18 have for every wired Wireless client so we have the state machine we merge the application data and again it's an optin

29:25 you as a network administrator have the full Authority to um authorize this

29:31 information exchange if you authorize it with the zoom and team credentials that's when the data comes into

29:37 your tenant in the Juniper Miss Cloud platform and all privacy is maintained

29:43 it is all data is all anonymized but the key benefit here is now you know proactively first of all how many Zoom

29:50 or team calls are happening within your network how many of them were successful this is again where it it all goes back

29:57 to the user experience we're able to zoom in on the bad user minutes for a

30:02 collaboration application and not only that we are then able to tell you

30:07 exactly where the issue is what capabilities of the network or the client or even van were probably

30:16 impacting that Endor experience what's the strongest correlation and that's where the explainability aspect comes in

30:22 so apart from having a wireless service level expectation metric per user or a wired service level expectation per user

30:28 user or van link uh SLE and application Health SLE we are now announcing an

30:34 application experience SLE especially for collaboration apps which comes in

30:39 you know especially in the Life Sciences model with your Telly Health that we are seeing a lot of now um this is critical

30:46 because you want to know that the users connecting to the network are having the best experience possible and if not

30:52 again Network it should know in their fingertips why that is with this prediction model we're able to not only

30:58 explain a bad Zoom call by identifying it but more importantly as you're rolling out collaboration apps in your

31:05 network we're also able to predict what the application

31:10 experience would be for voice or video right and here is again a quick demonstration where we've pulled in zoom

31:17 in teams data from uh the zoom in teams Cloud instances for your accounts into

31:24 the Marvis framework and we are now able to very quickly and proactively identify

31:30 bad user minutes and what is the reason why those user minutes are bad I'll now move on to Marvis minis

31:38 right everything I spoke to you before was all about client Telemetry coming

31:44 from our Network where we're building the state machine but often times the question

31:50 comes hey what about the time when I'm rolling out a new location I'm creating this brand new building or this creating

31:57 this brand new campus I'm going to deploy the network how can the network

32:02 assure me that I will have the best application and end user experience

32:07 possible before even the first client connects to the network I will now introduce you to

32:13 Marvis minis Marvis minis is the industry's first digital twin for user

32:21 experience we are now able to give you complete visibility into what the user

32:28 experience will be when a user comes in to connect to the network before there are any users on the network themselves

32:36 so here what we have done is essentially created because of our microsources

32:42 Cloud architecture we've created a cloud native minis framework again imagine we

32:48 talked about Marvis actions we talked about Marvis convers interface now you have Marvis watching over the network

32:55 and making sure that the end user experience is the best possible even before the first user connects to the

33:01 network it's always watching Always monitoring and always validating that

33:07 the services will be available when the users log on to the network proactively and that's how we call it the industry's

33:13 first digital twin for user experience the Maris Min framework is available on

33:18 the wireless as well as the wired and we also recently announced U on the van side so again putting it end to end and

33:26 most importantly it's able to test the availability reachability of the Network

33:32 Services as well as your applications and also it's able to predict what the

33:38 experience would be going back to the user learning model I just spoke to you about where we're looking at the loss

33:46 latency Jitter on the van side we're looking at any authentication issues on the wired side we're looking at all the

33:53 V lines being available all the applications being available for for the end users before they even connect to

33:59 the network with this Marvis Min framework but again the key problem statement was when I have clients

34:06 connected to the network I already get the Telemetry I need to be able to tell you what the user experience is but as

34:12 you roll out new sites how do you know that you'll have the same experience that you're seeing on other locations

34:18 without any uses connecting and that's where the marus minis framework comes into play it's always on it's always

34:26 validating the the network experience for the end users even before any clients are there so again as I

34:33 mentioned earlier it's available for both Wireless and wired and we recently just announced at an event um that we

34:40 did in June on the vanite as well so here you can see whether it's DHCP whether it's ARP whether it's DNS

34:47 whether it's application availability whether it's authentication on the wired side whether it's the speed test on the

34:53 vanite we are enabling this all dynamically there are there's nothing that you have to do or set up Marv is

34:59 running in this for you in the background all the time to make sure that from a connectivity perspective

35:05 from an authentication perspective from an application performance perspective the network is always available to

35:12 deliver on the application experience that we need for the end users to have

35:19 the best experience possible when they connect to the network right so the key part again is dynamic in nature it's

35:25 always running and it's able to validate for you what theer experience would be

35:30 proactively before the is connect it also has a role to play from

35:35 a brown field perspective so apart from new deployments how do we make sure the Network's always available how do we

35:41 make sure the experience is not impacted when you make configuration changes right or if by some reason

35:47 misconfiguration happens how can we cat that proactively Marvis Min is the answer

35:52 it's constantly learning your network it's constantly baselining the network it's constantly testing for availability

35:59 of the critical Network Services and applications no additional Hardware no

36:04 separate platform it's all part of our Juniper M Stack the same access point portfolio that's delivering these

36:11 capabilities for you and when Maris minis find an issue again you don't have

36:16 to look at the dashboard or ask Marvis hey is that a problem Marvis will

36:21 proactively notify you as part of a marus actions framework hey I found an issue I'm seeing latency law from this

36:28 application please expl to the application team or I found a DCP issue or I'm seeing a loss on this particular

36:35 internet circuit that's coming into uh your hospital or or Healthcare campus so

36:42 that you can take the appropriate action with the right teams to mitigate minimize remediate the problem right

36:49 away now we talked a lot about user experiences but the other key aspect is

36:56 what about on boarding devices and here I'll pivot towards our missed access Assurance solution right today there are

37:03 existing Solutions in the industry that enable secure on boarding be with be it the guest access use case be it the BYOD

37:09 use case your iot devices coming online um or even your critical corporate

37:15 devices with certificates coming online but the architecture that's available today to support that is extremely

37:22 monolithic it's extremely brittle it's extremely complex what we did did again by virtue

37:29 of our microsources Cloud architecture we essentially shattered that complexity so now again leveraging the simple Cloud

37:39 microsources framework we have launched since last year and it is now in

37:44 production for the last one year and we've done we have multiple customers who are using this um it's all about

37:52 leveraging secure on boarding with our access assurance

37:58 microservice no more appliances on site no more real estate that you have to

38:03 worry about to manage your a huge cluster of appliances it's a simple configuration in the cloud and more

38:10 importantly it's completely integrated with the same client visibility that I spoke about before from a wireless wired

38:17 van perspective so what when now when you us are you want to enable the secure authentication be it via you know ppsk

38:25 or be it via 1X or be it just a guest access solution you are fully assured of

38:32 the same Cloud platform delivering those sources for you at scale with a single

38:38 click of a button all you do when you set up a wireless land network is say hey authentication services use misto or

38:45 from even from a wired perspective for 1X authentication all you do when you do the setup is as part of the wired

38:51 Assurance configuration for templates at scale just put the authentication Service as use M off and you create a

38:57 very simple policy framework and now everything is available to you not just

39:02 connectivity not just the experience of users when they're connecting to Applications but as Matt brought up

39:10 before security is obviously a big big focus and parameter you are now assured

39:15 of the right devices connecting to the network in the right way at the right time accessing the right resources and

39:24 if those users are having an issue the uh the by virtue of this being in the

39:30 same Cloud platform as an additional microservice you have full visibility into that pre connection experience as

39:36 well whether it's a a bad password or an expired password or an expired

39:41 certificate now you have the full visibility in the access Assurance

39:47 module of the Juniper Miss stack in the cloud platform of knowing hey first of all which users are accessing Water

39:53 Resources in a very simple framework like you see here and more important if they're having a bad experience is it

39:59 because of authentication problem and what exactly is the problem it's right there for you at your fingertips again

40:05 the right data the right architecture at the right time but how do we make it even easier again when you're

40:13 troubleshooting a user and issues come in you want to be able to empower your level one right today what happens when

40:20 any network ticket comes in it goes straight to level three and the level one team say y i acknowledge you have a

40:25 collaboration issue problem a voice video problem or authentic or you're not able to connect to the network let me escalate to the level three what we have

40:33 done with the AIML framework and that this is what you should expect of your aops platform is leveraging a conversal

40:39 interface be able to very quickly troubleshoot complex problems but very

40:44 simply so now your level one help desks is a lot more productive they're not escalating every issue up to level three

40:51 and they're able to solve problems right away when a ticket comes in to know exactly what the issue is especially

40:57 it's tied to whether it's connectivity user performance application experience like Zoom or teams or even from an

41:03 authentication perspective again no more guesswork no more blind spots end to

41:08 endend visibility last but not least if you talk if you remember what I you know how I started my conversation was it's

41:15 all about the visibility end to endend it's all about the a driven support with Marvis it's all about the Microsoft's

41:22 architecture that makes it all possible but it's also about what else can we do

41:28 to enable engaging experiences for patients and visitors for the clinicians

41:34 for the nurses for the caregivers to ensure you're getting the best possible experience from the network and this is

41:41 where indoor location services comes in again part of the same Cloud platform not an overlay solution not a siloed

41:48 solution and no more overlay networks required so this is something again that you can deploy at scale it's part of the

41:54 same wi-fi system leveraging a standards based Wi-Fi and Bluetooth low energy technology to deliver for you the best

42:02 experience possible this is where again we're able to deliver asset visibility use cases

42:09 working with our ecosystem Partners we're able to deliver uh wave finding use cases with our ecosystem Partners we

42:16 essentially become the platform standards based technology on Wi-Fi and blle with differentiation built in so

42:23 all of our Flagship performance APS have this directional Bluetooth antenna array built in and the reason that's important

42:30 is it because gives you higher accuracy when you're looking to find an asset in room um you should be able to find that

42:37 very easily without deploying a proprietary stack or a silo system that's exactly where the jiper Miss

42:45 platform comes into play when you deploy a network you want to be able to leverage location services

42:52 part of the network offering and do that the day you deploy a network and that is

42:57 where again an AI op stack comes into play We leverage unsupervised machine

43:03 learning as part of our system to give you the best location services possible with the highest accuracy that

43:11 ble can provide converging with Wi-Fi last but not least all of this

43:19 also is available to you why what we' have done with integrating with a service now platform so the the

43:26 underlying fact for client visibility AI driven support um Microsoft cloud architecture

43:33 digital experiences also is all based on the entire platform being 100% API first

43:39 these apis help us integrating with partners for enabling the location services use cases these apis also help

43:46 from a network it perspective deploying the network and operating the network so every Mar action you saw can now be

43:53 leveraged as an auto ticket opening up in the now platform so if you have

43:58 service now as your key platform for it operation management this is where you

44:04 can Leverage The Integrations we already have for speeding up your deployments

44:10 with enabling the workflows wire service now interacting with our apis or even from a day two perspective leveraging

44:16 the service graph connector application we have built to deliver simplified day

44:21 to operations and make sure that these are part of your workflows that your it teams are already used to by leveraging

44:28 the best of breed of solutions with the aops combining that with the service now

44:35 solution and last but not least as part of our Enterprise uh blueprint we also

44:40 recently announced the ability for Marvis to now get information even from

44:46 anomal seen in the data center so for our abstra Control Data Centers Abri are

44:51 intent based networking solution for the data center we have now been able to

44:57 also build a full end to end view where if you have application sitting in the

45:02 data center that are having issues because of data center problems how can we very quickly from an endtoend user

45:09 perspective from the client to Cloud client to application give you that full view of if there's a problem if it's not

45:16 my wireless not my wired not my SD van but it's a data center how can your

45:22 Hospital campus Network operators know within the click of a button exactly where the problem is so the first thing

45:27 they get to now know is it's a data center problem secondly they can now enable Mars for the for the data center

45:32 and know exactly what the problem in the data center is that could be impacting Downstream users like your clinicians

45:38 like your nurses or even patients accessing applications and again getting to the view of what is my Ender

45:44 experience if it's bad why is it bad so proactively letting you know exactly what the problem is by pulling in data

45:51 center anomalies and Telemetry as well into a Mars framework engine so just to summarize all of that the zoom in themes

45:58 integration the integration with Z devices to full Telemetry to give you the full view of what's Happening how

46:03 does a client see the network how does a network to the client giving you a view of what the service level experience is

46:09 for every user wide or Wireless um the ability for us to securely onboard users

46:16 the right way at the right time with access to the right resources to ensure that your security postures maintained end to end the ability for us to deliver

46:23 indoor location services with our ecosystem Partners to make the are the possible come to life around asset visibility around wave finding around

46:30 push notifications and the ability for us to put an information even from now a

46:36 juniper AppStar data center into the Campus View so you know if an end user

46:41 impact is happening is it because of issues beyond your campus Network all of that is possible viable again because of

46:49 our foundation in an API based Microsoft architecture that is what allows us to

46:55 enable these services and with the agility that has not been seen in the industry so far but end of the day that

47:03 microservices architecture is what truly delivers the benefits leveraging aops

47:08 leveraging AIML AIML is not the end goal the goal is how have we been able to

47:15 leverage a IML to deliver iops to again deliver the best user experience

47:21 possible on the network as your devices connect on the network and continue to

47:26 use a network for the day-to-day services and with that Matt back over to you thanks unini I think everybody's

47:33 head is spinning so time for me to to bring us home so quick glance for those of you that may or may not know who we

47:39 serve right so standing behind our 5.5 billion in annual revenue you'll few see a few of segments that that are

47:45 mentioned here in particular uh based on what you've heard over the past half hour a couple of interesting facts so

47:51 we're the fastest growing Wi-Fi vendor in the market our campus business which soonly address very very nicely so think

47:58 wired and wireless it's the largest it's the fastest growing unit within the company we see uh rapid adoption across

48:04 all different Industries which you see listed right here including Healthcare where we've been growing at a double- digit clip annually so very very healthy

48:10 business pun intended and uh you know to gives you a little bit of an idea and a little bit of a peak on who we serve

48:17 right from a customer perspective next slide so uh what some of our customers

48:23 say about us just at a very very high level from radi children to AA Health to Island Health you'll start to see some

48:29 of the kpis that really kind of get extracted as soon said uptime you know typically is you know not good enough

48:35 anymore and slas and the metrics that are around Network up time are oftentimes obviously what we capture but

48:42 now we start to see really where AI where some of the simplistic nature of what we built in around automation can

48:49 be captured from a kpi perspective and the differences that can be made within these types of environments R Children's

48:55 Hospital perfect example deploying over 900 APS less than 6 weeks uh ztp Cloud

49:02 single click activation function and features right allow them to do this uh the important part here too Is Not only

49:08 was it fast but there was quality behind it so no no disruption zero trouble tickets have never experienced anything

49:13 like that before a direct quote that came from uh the chief technology officer uh to AA Health where now

49:20 because they've moved off of a controller based architecture to a more modern microservices Cloud architecture

49:26 they're saving over $1,400 a year not taking down and bringing back up those controllers right that have been a pain

49:31 point for a decade Plus in their environment uh which was equivalent to about 70% of an FTE to then Island

49:38 Health who's talked about basically allowing Marvis and allowing the technology to really inspect their

49:44 environment and seeing a dramatic drop so 90% drop in user generated trouble

49:50 tickets so just a few examples that are that are there for you all the next slide um Partners customers

49:58 prospects they put a lot of stock into the words of Industry analy so it' be remiss of me excuse me not to uh to

50:04 share this particular information on the le- hand side wir and wireless land magic quadrant it's a third year in a

50:10 row that we're in the very very top spot additionally we're also a leader in the indoor location services which is on the

50:15 right you're going to see that we are the only networking vendor that's up in that upper right hand quadrant and again

50:21 this is primarily due to again what I'll call is the Swiss army knife uh AP approach where we embed virtual

50:27 Bluetooth low energy we derive those location based services that sun Alini just talked about directly out of the

50:33 access points and then the next slide is one important item of note here I wanted to

50:40 put this in so the Gartner critical capabilities report some people pay attention to this um some people don't

50:46 know that it's available so I wanted to make sure that everybody was aware of this this one is for wir and wireless land you'll notice that Juniper achieved

50:53 the highest uh scores out of all vendors and all five out of five use cases uh

50:59 we're super proud of this accomplishment and honestly you know what this comes down to is really thinking our customers

51:05 uh especially for the feedback right that goes into these and driving the results out and so I'll wrap by just thanking

51:12 each and every one of you for joining us I hope the content was of value I know it was a lot of information uh please

51:18 reach out if we can be of assistance and we'd love to be part of your journey so all the best and thank you for for

51:25 attending thank you

Show more