Zero Touch Testing for an Assured Service Experience

0:02 hello everyone thank you for joining us today we're thrilled to be here because this is our first webinar since the

0:09 Juniper Networks acquisition of net rounds hopefully you've all heard by now that

0:14 netrons has been rebranded and is called Paragon active assurance a part of Juniper's Paragon automation

0:21 portfolio today's webinar will be held by Juniper Network's surreal desau product

0:27 marketing director for Paragon Wan automation previously the VP of marketing at that rounds

0:34 he will be joined by analyst Mason's Anil Rao principal analyst and

0:39 researcher today's agenda will start by identifying the need to improve customer experience

0:45 and satisfaction in Telecom engineering and operation teams the

0:50 challenges that they have and how automation can help them how to leverage active Assurance as part

0:56 of the full service life cycle the service activation testing problem and how you can easily solve it

1:03 key characteristics for automated active assurance and moving into a tier one success story

1:10 followed by finally a q a session now you will see at the bottom of your

1:16 screen a q a box please feel free to write your questions in that box during the

1:23 presentation some questions will be addressed verbally at the End by our speakers

1:29 alternatively you'll receive a direct reply in the chat box

1:34 without further Ado cereal over to you thank you Natasha my name is sir duso I

1:40 am a director of product marketing at Juniper Networks I'm quite happy to be co-presenting here with Anil today from

1:46 analysis Mason Anu Anil from many years and is truly an expert in service assurance and automation

1:53 I would also like to thank you everyone on this webcast and for participating to this webinar on zero touch testing

2:02 as we all know communication satisfire can differentiate with Superior customer experience

2:08 this is that not surprising that in the Telecom study from ads unearning that we

2:13 see here we can see that 78 percent of respondents claim that customer experience and satisfaction are top

2:20 priorities however as we can see one of the survey participants responded that currently

2:26 our industry is plagued with churn levels that will terrify any other Industries

2:32 so let's see how we can improve this and so when talking about customer

2:37 service experience there are many aspects that we could focus on for example we could make sure that we have

2:43 the right offer um that for the the bidding processes are easy and automated

2:49 in this webinar we're going to focus on explaining how a service provider can deliver the right service

2:55 characteristics from day one and how they can make sure that engineering teams can onboard customers

3:01 quickly without impacting customer experience all thanks to our zero touch

3:07 testing solution so one of the issues that our industry has had for many years is that customers

3:14 are first to find network problems and a recent survey from Juniper Networks with 200 large companies in

3:21 North America revealed that network problems are not reported or reported by

3:27 end user and only 39 of them are discovered by network operations

3:32 this means that more than 60 percent of network problems are discovered by customer are not reported at all

3:41 this is something that needs to change if we want to achieve our strategic objective of improving customer

3:46 experience and satisfaction so prior to describe the solution and it

3:51 will now explain what our communication service for your wider challenges and

3:57 how the process needs to evolve to assure customer experience

4:02 thank you Cyril um that's a great introduction to why

4:08 service providers must put customer experience at the heart of their business strategies

4:15 um especially in the context of network service experience so in my part of the presentation I will

4:22 argue that service providers need to think about a new operations model to to

4:30 address some of the challenges that Cyril explained in his initial remarks

4:36 and I believe Network automation is foundational to

4:41 this new operations model so I'm going to try and explain this in a little more detail

4:46 so in the chart here you see I've tried to sort of um depict the the traditional operations

4:52 model highly simplified of course so in this operations model

4:57 um the organization design and the business processes um were designed for traditional voice

5:04 and data services as siled and highly manual processes

5:10 so network engineering and field operations you know build out the network

5:16 followed by the business and I.T operations creating the customer orders and then it and network operations

5:23 provisioning the service orders and finally the operations and Engineering teams managing the day-to-day Network

5:30 and service operations into Department collaboration

5:36 um has been critical to make this model work but then this has been done largely in a

5:43 manual manner this is one of the key reasons why service providers face some of the

5:49 issues that Cyril explained earlier so the siled processes with manual

5:56 interdepartment handovers lead to an inward focus on operations losing sight

6:03 of the the broader business goals such as customer experience

6:08 Superior customer experience and so on so in my part I want to propose a new

6:14 operation model for what we call is the end-to-end service life cycle automation which is

6:21 fit for purpose for the 5G error and more specifically

6:26 um to achieve these broader goals including reducing Opex increasing service agility

6:34 and asil was saying delivering Superior customer experiences

6:40 also a new operations model will be necessary to support new business models

6:48 such as Network as a service or network slices and so on as well as managing the

6:53 network complexity of the sdn and cloud-based 5G networks that we are

7:00 starting to operationalize on an ongoing basis here is a simplified illustration of the

7:08 what what the industry calls as the intent driven Network automation framework

7:14 for service life cycle automation now one of the basic principles of this

7:21 new operation model is that the business intent drives

7:26 um the rest of the operations now the business intent could be an Enterprise uh wanting a low latency Network slice

7:34 or wanting to set up 100 sites with sd-wan as an example

7:42 so this business intent drives the automation of the service design

7:48 for that Network slice so at this point the service

7:54 characteristics of the network slice is encapsulated in the service

8:00 design which is then translated into service intent through the service orchestration

8:07 process the service orchestrator then orchestrates across one or more Network

8:14 orchestrators which which further encapsulates the sort of the functions

8:20 for translating the service intent to network intent

8:25 so during this process the network orchestrators instantiate and activate the the networking resources

8:32 associated with the service as well as validating and monitoring the network

8:39 and service um as well as performing analytics driven automated actions

8:47 um so these automated actions would then drive the the self-healing process of

8:53 the network um on the service um before they have an impact on

8:59 customer experience so at this point we've we've sort of moved from

9:06 uh the design time operations to the to the runtime operations and then back

9:11 into the um Handover back into the runtime operations uh with with the with the

9:19 um monitoring of the customer experience so in this mode uh the service life

9:25 cycle within runtime operations is a fully automated policy driven uh

9:32 self-controlling loop um with the necessary touch points with the design time operations with with the

9:39 Handover from the service orchestration to network orchestration on one hand and

9:44 the the self-healing and and the and the automation of the runtime operations back into the customer experience and

9:52 back into the business intent um so that's on the that's the other way

9:57 now according to our latest numbers um there are now more than 100 5G

10:04 networks launched in over 44 countries so

10:09 even as these Networks are being launched um service providers are telling us that

10:16 they are in the midst of a plethora of network automation initiatives

10:23 so in this survey of of 55 service providers that we did more than 51

10:30 percent uh said that Network automation is a top three strategic priority

10:36 and then that number you know increases to 75 percent uh who said that it is a

10:42 top five initiative so when you add the the the top two categories

10:47 now when we probed further service providers also stated that increasing

10:54 service agility um leading to enhanced customer experience and reducing Opex were the

11:02 key drivers for Network automation now achieving the goal of full Network

11:09 and service automation like I explained um in in in in in in in the chart a

11:15 couple of slides ago now achieving that sort of uh full automation

11:21 um will be a multi-year journey and it will require an incremental or a stepwise approach

11:28 so analysis Mason believes that service providers will need four key highly

11:34 inter-working OSS pillars to achieve this goal so we need on one hand the service

11:40 design and orchestration systems which enable automated Network planning design rollout and and service design functions

11:50 um supported by automated customer and service order orchestration right so this is mostly on the on the design time

11:57 operation side of things so these systems will will translate the the business intent into service intent

12:04 like I was explaining earlier and and Design Services by decomposing the

12:09 customer request which is the business intent into network service requirements

12:15 second Network orchestration systems uh will translate the service intent into

12:20 Network intent managing and and dictating the the network resource life cycles within as

12:28 well as across domains as well as as we're seeing

12:35 um with the introduction of edge clouds and and Central clouds and so on the

12:40 orchestration systems will have to work across these distributed Edge clouds as well as Central clouds to ensure that

12:46 the the service services are activated when and where required and configured

12:53 as needed third a robust Assurance platform or a

12:59 solution is going to be essential to support automation at every layer and

13:05 every domain of the network so the Assurance solution monitors the network and the service performance as well as

13:12 validating the services as they are activated um and and the solution must integrate

13:20 into higher level orchestrators to achieve the the closed loop automation that I was highlighting earlier

13:27 and finally fourth the mlai based analytics will provide the key insights required to drive

13:34 assurance and network and service automation as a whole so embedding ml AI

13:40 in all the layers of the operational stack will ensure that the execution of

13:47 mlai and the subsequent actions are are applied close to the point of need

13:54 next so with that backdrop of the network and

14:02 service lifecycle automation framework let us double click on one particular

14:07 aspect of automated Assurance which is the theme of today active testing

14:12 now active testing has been traditionally used among other things for test and turn up test testing of

14:20 services before handing over to the customer so if you remember

14:25 when I was explaining the you know current operations model this is typically a manual Handover

14:32 between the the it and network operations Department who are responsible for service

14:38 provisioning the hand over to the operations Department who then conduct the service

14:44 validation before handing it back to the customer facing departments um for you know ultimately hand over to

14:51 the customer now if we take that and then

14:56 sort of apply the new operations model this process can be fully integrated

15:04 into the network orchestration process as described here in the chart

15:10 so at the end of the provisioning process the network orchestrator can

15:15 automatically invoke the active testing routines to perform the the process of

15:21 test and turn up validation and the results of the test can be

15:26 automatically automatically communicated back to the orchestrator which can then further take actions such

15:33 as communicating with higher layer OSS such as service orchestration and hand over the service to the customer

15:42 so in addition to the test and turn up use case um active testing automation can also be

15:47 applied as I was saying to a plethora of other use cases and some of which are explained here

15:54 first the validation of network changes before release to the customers so this

16:01 is an extension to the to the test and turn up testing um and provide further benefits by sort

16:09 of significantly reducing the change management cycle time which can take days or weeks

16:15 let me run through the other use cases validating

16:20 um that the network supports business objectives now this could be relevant to a wide variety of scenarios as well so

16:28 but as an example it could be for proactively monitoring quality of service of a mobile backhaul Network or

16:36 The end-to-end Poisoned data service quality to deliver on the contracted

16:42 slas faster fault isolation so this is about

16:47 using automated active testing to test individual Network segments to isolate

16:54 the cause of performance degradation validation of cross-csp services so

17:00 we're seeing uh demand for these type of services so so this is about testing the services

17:07 that span multiple csb networks and and multiple clouds as well

17:13 validation of hybrid Services um it's in it's in the title it's it's about services that span both the

17:20 physical and virtual Networks and finally validation of vnfs and cnfs

17:27 and services that that span the the virtual Network so as an example this

17:33 could be about validating and certifying VNS and vnfs and cnfs the cloud native

17:39 Network functions um and the services as they are instantiated and activated

17:47 with those remarks um let me hand you back to zero cereal over to you

17:53 thank you Anil um so as Anil mentioned today we we really want to focus on service active

18:00 action testing and some of you actually might not know what every engineer needs to do when

18:05 they onboard a customer so I just wanted to provide an example of a typical

18:11 activation test uh testing so when we um

18:16 when a prior to onboard a customer you might need to basically double check that the security Gateway are well

18:23 configured you might need to double check the qi's profiling on the end-to-one service delivery chain

18:29 you might want as well to double check that specific service like a vote

18:34 service can run properly from unto one uh if you're offering a managed service

18:40 you may want to make sure so that you have which ability to the the managed service from your customer location and

18:47 if you're using Partners you might want to make sure that your partners are delivering upon their their SLA and if

18:54 one of those tests is not working like for example if you're doing a lot testing then you can't onboard the

19:00 customer yeah you have to make sure that everything is properly configured prior

19:06 to onboard your customer and in fact while you we we have to do

19:12 all those tests in reality services are not tested sorry and and I think I

19:18 didn't mention it a bit as well and the reason is that there's um we used to use a lot of handle test

19:25 device um there that they might be difficult to to to test and so at the end of the day

19:31 to use to test those surveys at the end of the day most of the team they they also pressured by by the business so

19:38 they do nothing they might do a simple ping and if they have a high value customer then they might send some

19:45 Engineers on site to make sure that things will work for that high value customer yet that is quite expensive

19:52 so at the end of the day if they are not testing soloid their services they have failed deliveries which leads

20:00 to bad reputation churn and on top of these very expensive repairs because when it doesn't work it's really tricky

20:07 to understand why it doesn't work so the solution that we're we're proposing today and that we are talking

20:13 about is using active testing which are synthetic transactions that we can run

20:19 on the data plane and by running those synthetic transaction on the data plane the the

20:24 way to do it is we are leveraging some software only virtual test agents that

20:32 can be deployed in the network and from those agents you will run the synthetic transactions and to one or

20:40 between the different network segment that support your service and that is

20:46 really the only way to properly understand service quality as an end

20:52 user if you don't want uh or cannot deploy

20:57 agents everywhere then you can use reflector technology in your network to

21:04 do some testing from the agent to a reflector so there are different

21:10 capabilities and possibilities depending on your service and deporting on your network Arctic architecture

21:16 but by doing this you are really going to understand your Antoine performance or Antoine service performance for the

21:23 characteristic that you sold to your customer for example you will understand if your HD1 service works and to one and if it

21:30 doesn't which is the under relay Network that might be damaging that performance

21:36 you will be able to to make sure that the Qs is configured in all the different network segments that are

21:43 supporting that service if there is a firewall that is misconfigured yes as well you will be

21:49 able to spot that basically your customer is not going to be able to reach that critical application when the

21:56 service is trying to go through that firewall and so there are a number of key success

22:02 factors that are required in order for you to succeed with those projects

22:08 one of them is to have one platform that offer all the tests that you need from

22:14 layer 2 to layer 7. so if you're offering a layer to a layer 3 VPN

22:20 service if you're offering an sd1 service or if you offering a web service

22:25 or a manager of Office 365 or any other manage application you want to make sure

22:31 that that platform that you have give you the ability to test all those applications and then when you deploy

22:37 those software only agents in your network that by the way can be automatically deployed basically you

22:44 have all what it takes to monitor your service and when others are critical success factor

22:51 is that you have the ability to build um through test template to compose the

22:59 tests that you need that depends on your service and that depends on your customer so for each of your services

23:05 there's a number of tests that you want to be done and there's an a number of uh

23:10 thresholds that you want to set that will be based on your basic

23:16 um I would say service definition or it could be tailored to each of your customer if you've sold a specific

23:22 takeout service to a customer you also want to be able to test the characteristic of that specific tailored

23:29 service and so you need that solution that gives you the ability to compose

23:35 um to to select all the tests that you need chain them and run them once but if

23:41 you have um different type of services like a web service on top of testing the

23:47 connectivity you might want to run in order to make sure that the verb service is is running you might want

23:54 also to try to reach your Office 365 service from the customer location to

24:00 make sure that the employees that are going to use that service are going to be able to get good performance about

24:07 all those applications that are hosting in the cloud and somewhere else and so I'm gonna head over back to to Anil who

24:14 has some some more critical success factor to share with you and so I need

24:20 back to you thanks Cyril so so when you sort of

24:27 um contextualize uh what Cyril was saying just now and and my initial remarks

24:34 um around the the whole network Automation and and the role of active testing automation

24:41 um so we need to now uh think about the

24:46 underlying technology that underpins the active testing solution right so Cyril

24:52 alluded to it in one of his slides um where he was talking about virtual testations or software only test agents

25:00 right so that's that's that's interesting because what I'm going to talk about is is something on those

25:06 lines here so um one of the additional requirement is

25:12 is that the the solution should not only support existing physical networks which

25:17 is most of what we have today um but should also be future proof to support the the new software defined

25:24 networks and virtual networks of the future so when we analyze from from these

25:31 standpoints um we will realize that the existing active testing technology which

25:38 which are predominantly Appliance based um may not be more suited for these uh new

25:45 requirements so I'm proposing here that the active testing technology must evolve to

25:52 support new deployment options um which sort of highlighted here in the

25:59 in the right side chart to start with the the the the commercial Hardware based uh deployment model so

26:06 this provides the the flexibility for service providers to deploy it on

26:12 their own preferred Hardware um and they can procure the active

26:18 testing um uh solution as a software only license so it gives them

26:25 um better flexibility vnf so this is the the next uh I suppose

26:32 both as an evolution as well as an additional deployment model

26:37 um is is is the option in which the active test components are deployed as

26:43 virtual functions um and and by Nature the virtual

26:48 functions enable a much higher dynamicity and and programmability

26:55 so so we're getting more and more flexibility in in the testing solution itself and finally

27:01 um containerized deployment model is in which the active testing components are deployed as microservices

27:09 and again by nature microservices enable a much granular control to scale and

27:16 upgrade um only those components that required so you don't have to uh

27:23 apply the same actions for the entire test platform so you could pick and choose the components that that you want

27:30 modified and so on so the idea here is to have a portfolio of deployment

27:37 options so service providers can can make choices as they deem fit for their

27:45 purpose and depending on where they are in their evolution uh journey

27:51 in terms of Automation and so on so in summary uh I just wanted to highlight

27:57 that as you move along this Evolution for for the for the test solution

28:05 um there's going to be a higher level of control and ability to do much better

28:12 um automation um because of that granular control and the dynamicity that the the components

28:18 provide so it it just sort of is a fit for purpose uh to uh uh

28:25 to the broader automation goal that the operators want to achieve but at the same time supporting a broad variety of

28:33 deployment models including physical because many many networks are still physical and giving operators choice to

28:40 make their make the evolution based on their timelines is going to be critical

28:45 um as a success factor so to bring us back to the automation

28:50 framework I discussed earlier um the test automation solution would look

28:56 something like this that that we've depicted here in the chart so notice that the test orchestrator as

29:03 a new component that we've introduced here in this architecture the test orchestrator will manage the

29:10 full portfolio of of test components that uh highlighted in the previous slide so the idea is to abstract the the

29:20 test infrastructure from the higher layer OSS systems and in the process

29:26 provide a clean API based interface for easy integration and ultimately uh

29:34 better automation so the network orchestrator would consume this API to interact with the

29:42 test orchestrator for automating the the test and validation processes uh that

29:49 Cyril presented in detail and some of the use cases I highlighted in my part earlier

29:55 so at this stage um cereal I will hand you back for for your

30:01 final section and your remarks thank you and so

30:06 I'm going to start by just summarizing what we've uh what we've said when we want to deploy zero touch

30:14 testing that's something by the way that can be done quite easily without some key success factor or the first one is

30:21 you really need the platform with deep networking capability layer 2 to layer 7

30:26 traffic generation the ability to have flexible test template you need the ability to have a solution

30:33 that can be deployed quite easily so we didn't talk too much about a SAS model or an on-prem model but if you have the

30:39 choice that gives you the ability to either go full uh on on on on on a SAS

30:45 the project solution in SAS or if you prefer to do it to do it on-prem you can do it or you can you can start with us

30:51 and then migrate your your deployment to an on-prem but what is critical is that

30:56 what I need just mentioned that ability to have options for the deployment of

31:02 the agents with a software version VNA version and containerized versions

31:07 and then that platform has to be built for automation with rest apis net confined young models

31:14 and that is actually what we have with Paragon active assurance and with with bargain activations we can automate

31:20 activation testing provide real quality monitoring to operation and as well give

31:26 the ability to do very uh fast troubleshooting and even remote troubleshooting when needs to be and so

31:33 I'm gonna now talk about a success story so one of our team One customer

31:38 as we mentioned you know he has a traditional layer 2 layer 3 HD one

31:45 business service and they're they're operating globally and as such they have

31:50 a lot of Partners Network that where they are buying connectivity to offer their own services

31:57 and so they've used those software uh agents they've deployed them in some

32:03 Telco Cloud so in four Tech clock clouds all around the world to be to do all the

32:08 testing from the Telco Cloud so the testing could be done at the edge but in that case they choose to do it centrally

32:15 and by doing this they were really able to really understand prior thrown about their customers they were able to

32:23 understand are we going to meet our service level can we onboard the customer

32:28 and when that was not possible or when when the the identified that there was issues the the slas are not yet met

32:36 basically they're able to really troubleshoot very quickly issues on the Antoine Service delivery chain

32:43 and often and unfortunately but often some of the problems came from partner Network so they're able to look at the

32:50 issue in the partner Network and as well they're able to show the partner Network their reports explaining why they

32:57 believe that the program is in in the their Network and give the ability for

33:02 the partner Network therefore to resolve the problem quickly and so at the end of the day prior to

33:10 use the solution it took them five when they had an issue it usually took them five days to look at that issue

33:16 and five days so it's five days where you can't build your customer and and people are waiting everyone is

33:22 frustrating your account manager within within the organization is is pinging you trying to understand when can we

33:30 onboard that customer and with the solution now we take them two hours to

33:35 locate that issues and if that issue is located in a partner Network they have

33:40 the reports that they need to communicate with the partner and accelerate fixing those problems and

33:48 which means that they are now able to really onboard customer much faster

33:54 therefore also build customer much faster they've been able to reduce their service delivery failure rate

34:01 significantly and they're also able to able to reduce their Opex but what they explained to us is it's really the

34:08 benefits for them is that there was a lot of people doing some manual tasks that are not really useful and now those

34:15 people we save a lot of time for those those teams that I can refocus on more

34:21 critical issues they can actually optimize further the quality of the network and services because now they

34:27 have the time and all those tasks that they had to do to identify the problem and to fix those problems they don't

34:34 have to do this anymore so it's a so it's it's a great win-win for everyone within that organization

34:40 now activations is part of the Paragon portfolio and Paragon is a model of Excellence which is what our customer

34:47 can achieve with this portfolio of solution on one automation and as we discussed today automations

34:54 paved the way for an Azure service experience I will not have the time to highlight

35:00 all of the product that are part of the Paragon sewage portfolio and if you are interested we will be able to do up some

35:06 follow-up webcast but you can find a lot of information on our on our

35:12 [Music] website anyway so

35:17 who is Fargo you can plan model and forecast your service

35:23 you can orchestrate service activation you can automate service Assurance as we

35:30 we really have seen today you can leverage machine learning to

35:35 gain actionable Insight on your service and networks and last but not least you can optimize

35:42 your service infrastructure through closed loop automation

35:48 so that is the um the end of the webcast part of that that webinar and I would

35:56 like to thank you all for being here with us and for your attention I really hope that session was informative and

36:02 we're gonna now start our q a uh a thank you from my side as well um thanks Cyril

36:08 always a pleasure to co-present with you and uh and and back to the audience for

36:14 for some engagement and some interesting questions I hope so we are actually have a few questions

36:20 coming here uh and the first one I think is going to be for you and Ill it's um how critical is it to measure unuser

36:28 quality oh thanks Phil um

36:33 well I I think uh you know you can't really fix something

36:40 um that you haven't measured right so um right off the bat I would say that

36:45 um measuring and user quality is absolutely critical if not for anything uh purely because

36:52 um you know from a from a customer's perspective that's what they're paying the operators the service providers for

36:58 so it's it's the service level agreements it's the quality of service it's latency you know all the rest of it

37:04 uh you know in the context of um service quality so you know without knowing how bad uh

37:13 or how good the service quality is you can't really take action to fix it so

37:18 from that perspective Alone um it's it's absolutely critical to measure and user quality and secondly

37:26 um if you are in a position um to measure and user quality and you

37:32 have the the necessary insights and the data on on the existing end user quality

37:38 that the operators are delivering to the customers then you are in a much better position to to drive the whole closed

37:45 loop automation um processes that that we explained uh during the webcast so so you measure the

37:52 end user quality so you can you can take some action to fix it and improve the

37:58 service performance and the service quality but then you also use the insights to to sort of Drive some of the

38:05 um mlai based um closed loop automation processes

38:11 so thanks Joe um I see another question just popped up here maybe maybe I think you're you're

38:16 in a good position because you talk to customers all the time so one of the audience wants to know

38:24 um do you believe that open apis are now mature enough

38:29 to automatically turn on service quality measurement for each service or location

38:37 that's a good question and and definitively so with solution like Paragon active

38:44 Assurance a containerized software agent can be automatically deployed in the

38:49 network so tests can be automatically turn up

38:54 and executed prior to onboard customers also true service level monitoring can

39:01 automatically be tailored to the service characteristic purchased by customer as we we discussed doing doing that that

39:09 webinar and so this is all due to the maturity

39:15 of a solution like Paragon activations that can be driven from rest apis and we

39:20 have now many implementation with various csps and very some some quite

39:27 interesting one but the thing is service Assurance has been overlooked in the past due to the

39:33 complexity associated with deploying and configuring Solutions and and here what we are bringing with

39:41 Ashanti's a solution that is that is actually extremely easy to automate to

39:46 make sure uh your services will be assured properly from the one

39:52 so definitively we we are really mature and and basically that's something that

39:58 I think every service provider should do so I'm looking at those of questions but

40:04 actually I think we're running a bit short of time so what we'll do with Anil will we will stay on the webcast

40:12 and we will respond to the question on the text box and so for the other question there's

40:19 there's a number of them we'll we'll send back uh the answers to everyone who's participated to to that webinar so

40:27 don't hesitate to leave your questions and we will come back to you and so

40:33 thanks everyone for attending that webinar it was a pleasure to be here with you today and we we hope to hear

40:40 from you and and to interact in the future on that uh very interesting topic and thanks a lot Daniel to to for being

40:47 here with us today well thanks for having me through always a pleasure to co-present with you

40:53 um indeed automation is a is a great topic um and and we we we research this topic

41:00 with analysis Mason in a lot of depth and uh hopefully we'll have more interesting conversations in the future

41:06 as well thank you