AI-Enabled, Cloud-Delivered Network Automation

0:00 [Music]

0:03 my name is Kevin Landry and I'm a

0:06 product marketing manager at Juniper

0:08 Networks for the Paragon automation as a

0:10 service and today I'm at Mobile World

0:13 Congress and I'm here to show you a demo

0:15 on our AI enabled Cloud delivered

0:18 Network automation solution at Juniper

0:21 we're driven by experience and we care

0:24 to deliver amazing experiences for our

0:26 customers and their end users but not

0:29 only that we care about delivering

0:31 amazing experiences for our Network

0:33 operator customers that are running the

0:35 network as well so we're starting with

Cloud Metro

0:37 our Cloud Metro solution where we're

0:40 introducing our Paragon automation as a

0:42 service and we're doing this to produce

0:44 desired business outcomes for our end

0:47 users faster and be able to reduce the

0:49 mean time to resolution for problems and

0:52 be able to prevent them so the first

0:54 aspect of resolution that's a benefit is

0:56 its use case based the reason why that's

0:58 important is because this allows our

1:01 customers to deliver what requirements

1:03 they need and where they need it at

1:05 their own priority secondly giving

1:07 operations an easy button to be able to

1:10 simplify our user interface so that even

1:13 operators that have less experience are

1:16 able to use it we've reimagined the

1:18 design to be incredibly simple and

1:21 thirdly it's AI enabled so there's two

1:24 points on this that I wanted to talk to

1:26 you about so the first is we're able to

1:28 reduce all the noise many times Network

1:31 operators are faced with an overwhelming

1:34 sea of data and alarms to be able to

1:36 troubleshoot so what we're doing is

1:38 we're using machine learning and AI to

1:41 be able to reduce all that noise to

1:43 filter in and hone in on the root causes

1:46 of problems and so that they know the

1:48 Urgent actions that they need to take

1:50 and secondly we have ai enabled Trend

1:53 analysis and Predictive Analytics that

1:55 are going to be able to detect problems

1:57 that previously might not have been

1:58 humanly possible to find issues such as

2:01 gray failure analysis where you might be

2:04 able to detect a bad optical cable

2:06 that's intermittently dropping packets

2:09 and degrading performance or maybe a bad

2:12 temperature on one of our network

2:15 interface modules as well can produce

2:17 something like that so at Mobile World

Use Cases

2:19 Congress today we're demoing three key

2:22 use cases for a paragon automation as a

2:24 service solution the first is device

2:26 onboarding which allows you to deliver

2:29 Services faster a faster time to Market

2:32 because you're able to commission

2:34 devices really simply and easily and

2:37 faster the second is our Network

2:40 observability application that allows

2:42 you to have unprecedented insights and

2:45 visibility into the network and services

2:48 the third is our Network trust

2:51 application that allows you to establish

2:54 trust and integrity in the network and

2:58 also understand compliance deviation

Network Onboarding

3:01 so first we're gonna explore the network

3:04 onboarding use case here you're going to

3:06 be able to see a field engineer on site

3:09 deploying and commissioning an ACX

3:12 device now as you see in the video the

3:14 field engineer starts by cabling in a

3:17 Management Port and she's able to log in

3:19 to her smartphone app and scan a QR code

3:22 on the device to be able to bring up the

3:25 application which presents her with a

3:27 step-by-step guide of how to commission

3:29 this device it Guides Through every SFP

3:32 to be able to plug in as well as every

3:35 cable to be able to plug in and where

3:37 then remote connectivity tests and

3:40 Hardware Health tests are performed as

3:43 well as to see if the interfaces are up

3:45 Network performance tests also test

3:47 connectivity as well to ensure that

3:49 everything is up ready and running into

3:51 the slas so the network operations

3:54 center is immediately in sync with all

3:57 the work and the work across all the

3:59 field Engineers is supervised by by that

4:01 network operations center the network

4:03 field engineer is then able to upload

4:05 pictures and complete the work order so

Live Product GUI

4:08 as part of our Network device onboarding

4:10 use case let me show you the live

4:12 product GUI for the network operator so

4:15 in this GUI you can see how all the

4:17 noise of the many alarms that you would

4:19 normally have to contend with are

4:21 filtered at and you're basically able to

4:23 hone in on the Urgent actions that you

4:25 need to take so drilling into this first

4:28 Urgent action on this ACX 7024 we're

4:32 able to actually take a look at the

4:34 hardware and see what the Urgent actions

4:36 are that are needed you're able to see

4:38 things like the compliance advisories

4:41 you're able to get a view of the actual

4:45 device that's being seen and actually

4:47 understand all the alerts and issues

4:50 with ports you're then able to go past

4:54 this health view of the device itself

4:56 and explore things like interfaces

4:58 software for instance in the software

5:00 tab it's interesting that you're able to

5:03 take a look at the network compliance in

5:06 terms of the end of life of the software

5:08 whether that's an issue or a cert

5:10 advisory so this is our Network trust

5:12 use case in addition to that we're able

5:14 to look at the recent configuration of

5:16 the device so for instance here we could

5:19 see that there was an active version

5:21 that was recently committed and there's

5:23 other versions as well and you could

5:25 explore to see if there was a change

5:27 that maybe introduced issues by

5:29 performing a diff comparison so here you

5:32 can see a side-by-side configuration

5:34 comparison of this device in addition

5:36 you're able to see a compliance score in

5:39 this case it's a score out of a hundred

5:40 that we've calculated and it's pretty

5:42 low so let's investigate that further to

5:45 see what's going on here drilling into

5:48 our Network trust application you're

5:50 able to take a look at all the different

5:52 compliance issues that we've seen all

5:55 these are benchmarks that have failed in

5:57 this case contributing to that lower

5:58 score and let's take a look at one of

6:00 these just to see more more details so

6:02 here you're able to see a description of

6:04 the compliance rule that failed as well

6:07 as gain a view into how to fix that

6:09 issue so when you scroll down you're

6:12 able to see a fixed information

6:14 description as well as the exact CLI

6:18 commands that you need to issue down to

6:20 the network to resolve the problem

6:23 moving on to see more about our Network

6:26 trust view we're able to actually from a

6:29 business level perspective have an

6:31 executive level view of the trend of the

6:34 trust score and this helps you see if

6:36 you're actually getting more trustworthy

6:38 or less trustworthy globally across your

6:41 network as well as understand what your

6:43 highest device compliance score is or

6:46 your worst or the average moving on to

6:48 our Network observability application

6:51 you're able to see a view a high level

6:54 of view that Aggregates all the data

6:57 into simple dashboards that you're able

7:00 to hone in on what the key issues are

7:02 and let me explore this further through

7:04 another application that I have that

7:06 gives you a view across devices the

7:10 network as well as interfaces and Trust

7:13 help now in this view of this dashboard

7:17 you're able to see and drill down into

7:19 things like the device health and gain a

7:22 timeline view in this case there's no

7:25 problem with Device Health Network

7:27 Health but there is a problem flagged

7:29 for interface of let's jump into that to

7:31 see more details so we can see in this

7:34 case that most of the interface issues

7:37 are caused by link failures there's 988

7:40 interfaces affected so let's drill into

7:42 that to see even more detail of why

7:44 you're able to gain an even more

7:47 detailed View and find out that out of

7:49 the link failures half of them are

7:51 caused by link state issues and half of

7:53 them by interface temperature issues

7:55 you're able to get a distribution view

7:57 graphically of what's going on as well

8:00 as understand which OSS versions are

8:03 being affected and in this case you can

8:06 see that it's actually a older OS that's

8:09 causing the issue so perhaps maybe

8:10 upgrading to a newer OS might fix some

8:13 of the problems you're able to also see

8:15 a view of the models as well as get a

8:17 timeline view and understand the

8:19 effective items in terms of devices

8:22 interfaces and services and that's the

8:25 live demo for Paragon automation as a

8:28 service for the three use cases we

8:29 mentioned today so in summary Paragon

Summary

8:32 automate service delivers services with

8:35 a faster time to Market because we're

8:37 taking on a cloud delivery approach

8:39 that's able to produce significant

8:41 results

8:42 we're able to produce 70 faster use case

8:46 deployment compared to do-it-yourself

8:48 models that give you 40 cost reduction

8:51 on deploying the use cases as well as

8:53 maintenance in addition we're enabling

8:56 AI for many event benefits to be able to

8:59 reduce mean time to no and mean time to

9:01 resolve all in all Paragon automation

9:04 delivers you the business outcomes that

9:07 you desire for your network to produce

9:10 that faster times Market

9:12 [Music]