Southbound_ 3G_4G Scenarios, Network Integrations and charging functions(PCRF ,CF,CGF,ABMF) (2024-06-14 10_33 GMT-4)-en-US.docx

Southbound_ 3G_4G Scenarios, Network Integrations and charging functions(PCRF ,CF,CGF,ABMF) (2024-06-14 10_33 GMT-4)

0:00
Right, and, and also Liliana is, it was in my team, but he, she's in the, in the product, you know, actually she's a product owner.

0:08
She's in the product office right now.

0:10
So I think for some in the future for the knowledge transfer, I, I, I prefer like you go through Liliana first.

0:19
And if you know, because Liliana knows some like architecture as well.

0:22
If, if, yeah, if Liliana cannot answer the question, then maybe she can come to in Elton, not Naomi.

0:30
So, so you know, Liliana, are you OK with that?

0:33
Yeah, Yeah, it is.

0:34
OK, OK.

0:35
Yeah, OK.

0:35
Mm hmm.

0:36
Go ahead.

0:37
Yeah, yeah, yeah, yeah, yeah.

0:39
Elton, we are agreed that Alton will present high level design for 3G-4G networks for data SMSMMS, everything what we have it is on the high level and Vinay will present for the five GSA data.

0:53
It is completely separate system, separate web clouds and it is a start.

1:00
After that you can we will share those access to those documents, this recording and after that you will let us know if you need to have additional sessions.

1:09
Thank you, Liliana, I appreciate it and thank you Ben.

1:12
I would say Liliana, you know Saki and Nikhil have created a folder called product.

1:19
So I do now have access to the product folder, so you can feel free to upload the documentation there.

1:27
And I will, you know, I will take care of sharing that with the team.

1:33
OK, thank you.

1:34
OK then Elton, can you please start with the 3G-4G high level design?

1:47
So I will, yeah, yeah, yeah.

1:49
I mean my very slow Oh, you can see my screen right?

2:02
Yes, we can see your screen Elton.

2:05
OK, OK, thanks.

2:08
This is sure it's the recording started OK for for 3G-4G is for the southbound.

2:25
So I use this diagram to show the self wonder what we are we are bonding to I think for the postpaid here what applicable is the P gateway MMSC and SMSC.

2:40
But for the post usage at least I'm talking about from the mediation side on the on the rating side that there will be CCS, right The new one that we're gonna as part of this now the transport that we're going to implement and it's the CCLS come here.

2:58
We use a different picture to show the how the CCS is to be added to the listing solution.

3:07
I think for CCS is also following similar approach.

3:11
The CCS will generate the EDRTDR kind of sale to the offline campsite for OPS southbound, except the CCHCH F1, the five G1 that have been able to show a different separate picture.

3:32
We need to handle basically 3G-4G data, keep it away and the MMS message and SMS message here also we have a kind of exception is that for US roaming due to accent the roaming issue, we cannot the charge from the real time.

3:58
We can we, we received a signaling, but we cannot charge the SMS directly.

4:03
We have kind of offline we that we use the tapping recorder.

4:09
So this I want to mention that because of separate interface, so if the the US roaming SMS, we have it to charge based on the tapping.

4:18
That's what's happening today and for the the long term, I'm not sure it's part of this will change, but that's what we are doing today.

4:26
So if we so in that case it is we detected the US roaming SMS, we would just run 1/4.

4:33
So we will wait for the tapping to come in.

4:36
For postpaid, we also need support modification.

4:41
You know what, I like the better picture.

4:50
Elton Suhas has a question.

4:53
OK, go ahead.

4:54
Sorry.

4:55
Go ahead.

4:56
Yeah.

4:56
Hey, sorry Elton.

4:57
Elton, regarding the roaming, the one you mentioned, is it only applicable for SMS outbound roaming in USA or it's applicable for other event and voice only for SSMS?

5:11
So only for SMS.

5:12
OK.

5:13
Yeah, for for the data we, we are doing our accent roaming solution.

5:17
It's not launched yet.

5:19
That's for data to based on the US border sales side.

5:23
It's under testing that's not launched yet for just a second scope of this project CCS postpaid, it is only data SMSMMMS, Voltivoice and HSBA voice has been rated by another platform and it is not rating by sub CC postpaid voice.

5:45
And currently this is not part of the scope.

5:48
Voice is a not part of this current project.

5:53
OK, so domestic voice is also not part of the project.

5:55
Ileana domestic postpaid voice is also not part of the project.

6:00
Data are the project all kind of data 3G-4G5G roaming everything and also five GSA data is part of this project.

6:13
Voice it is going to be separate because the voice has not been for postpaid customers, voice has not been rated by sub CC goal of this project is to sunset sub CC rating system right?

6:25
So does that mean online mediation will route that events to a different system altogether at the moment and not to say sub CC for voice, we do have a little bit of postpaid, but the for the voice is the RF interface.

6:45
Then this the RF, the CGF is we would that be part of these part part act or we only do that we focus on this online charging and the CCS.

7:02
So what I'll say your question, the is it go to the postpaid voice?

7:07
Is it go to the the offline charging interface and the sync to the offline and the targeting but for prepaid, yes, online is doing that and similar to different rating engine.

7:24
OK, thanks.

7:26
Yeah, no problem.

7:34
Too many, too many window Yeah.

7:40
So for notification, so with what I'm focus on the online Android part.

7:47
Once CCS or sub CC today they detect or we need to send our notification that notification out they will see sub CC will be replying the response or you need to send out a notification.

8:03
The notification could be a block, it could be a threshold of notification like.

8:08
This percentage means they used a certain percentage of your bucket and or block the We directly send the message to to the SMSC with SMTP interface too, because that cannot be delayed, we never triggered.

8:30
We need to send out the customer receive, we call it a block SMS and they can either do a top up of and they can reply some keyword to unblock.

8:43
So that's the notification, the notification platform.

8:48
So it's not a southbound.

8:50
It's it's it's the the application that online they need to talk to another case is that another blog about the floating.

9:01
So let's see for some unlimited plan.

9:05
You wouldn't be blocked, but I once you reach use the oil bucket and aligns that we need to send the throttling to PCF by PTCF to to apply the throttling rule.

9:22
So this part of their the two solution.

9:24
I think for CCS, the idea is that we will go to the the ice wide based today's a super API whenever the throttle happens, we need to send the the Super required to PCR and to trigger problem.

9:40
But for tomorrow is that you SY based Mahesh, right.

9:47
So you have a question.

9:49
Hi, Yeah, right now you explain about the policy control routing to PCRF basically, yeah, Yeah.

9:58
So I understand now for prepaid already the metrics CCS is integrated with online MZ and also the PCF.

10:07
So for currently for prepaid already the SY and the North 28 interfaces are used for policy control or not?

10:16
No, not, not.

10:17
They don't have throttling price plans and this is the problem which actually we have about.

10:23
I would like to mention we currently not using for integration as by and and 28 interfaces and we do not have a tools to test those interfaces.

10:36
And as a part of this project, it was actually planned to create those tools and our understanding was that actually prod up team will create tools for testing SYN 28 and N40.

10:53
Yeah, definitely.

10:55
I mean, I just wanted to understand whether we were not pretty sure how it is for prepaid today, whether it is they don't have throttling, they don't prepaid doesn't have throttling, price plan doesn't have roming.

11:08
Yeah.

11:08
Now it's clear, Yeah.

11:10
So as you explained, now it becomes clear.

11:12
So basically as I understand that SY and N 28 will be a complete new integration as part of the Pushpaid project and we should test.

11:24
Yeah, understand.

11:25
Yeah, yes.

11:25
And we don't have test tool to test, We cannot have functional testing for integration testing for performance testing.

11:32
We don't have those tools for those interfaces and those interfaces tools has to be built as a part of this project.

11:39
And this integration it is is completely new functions will require a little bit more effort and time actually to complete.

11:49
Noted.

11:50
Noted.

11:51
Another question, but I understand the the P gateway is already triggering the GY or the N40 sessions for prepaid, right?

11:59
Or is it not the case prepaid customers do also do not have 5 GSA.

12:05
Ah OK.

12:06
They don't have 5 GSA.

12:08
It's only the 4D then Yeah.

12:10
And they don't have roaming, they don't have 5 GSA, they don't have throttling price plans, scope of prepaid is much, much smaller than a postpaid.

12:21
Ah, OK.

12:22
So the five GSA interfaces will be a complete new testing for us.

12:26
Then for portal product, Understand we are currently using for postpaid customers, but we are not talking with those interfaces with sub CC.

12:37
With sub CC we have specific API which we are calling.

12:41
We are not using N40 interface to communicate it to sub CC, but we will use N40 to communicate to CCS in the future, right, Yeah.

12:54
And and as a part of this project, yeah, I mean, I mean the current project postpaid project, yes, yes, yeah.

13:01
So the the the background is that sub CC the current charging engine is not a how to say it's not designed for telecom, right.

13:11
It does not support the full telecom stack.

13:14
CCS they support to support support all these 3GP standard interface means GYSY in 40 into date.

13:25
So and also that the guideline to to to the development team is there for CCS.

13:32
We will change the current customer's interface to a standard interface.

13:38
So that's why that's why it comes in right.

13:40
So we do have a throttling today without a CCS online similar super request is API to PCs, PCs will do the throttle, but tomorrow, yeah, OK, yeah.

13:57
So this this the CCS case.

14:00
And for that we do have some function already in place like all these CCS load of balancing and to to talk to different CCS and what we call it white list confirming the number.

14:18
Should it go to sub CC or should it go to CCS?

14:21
I understand that.

14:23
I my feeling that it is a kind of a unique function we could reuse for this postpaid migration.

14:30
But there's some unique case that like for the notification I mentioned in case of CCS, we have a REST interface on Amazon has a notification workflow.

14:43
So whenever CCS detect what a kind of notification in the send out CCS, we will talk to online that is a notification workflow and the notification workflow we will talk to the notification platform.

14:58
And you may wonder why CCS cannot talk to the platform or why we have these, these workflows because some of them that notification like the one I mentioned blog has to send out in real time.

15:16
And the notification platform, they have logic to, to send it to emails, send it to the account the owner, the customer has suppressed or figure what kind of notification they are interested.

15:29
So that's why there's a notification platform.

15:32
But this interface between the different workflow, today's platform engine is arrested.

15:39
It's a REST interface we already have, but a part of this postpaid migration for sure.

15:48
They they're going to be more kind of a typo.

15:51
The fiction will be introduced from the current sub CC interface to CCS.

15:58
OK, Elton.

15:58
And one, one more question, sorry, here in this presentation, in this diagram which you're showing, Elton, I don't see the N 28 flow.

16:10
So the as I mentioned, they this for the 3G-4G you will see.

16:15
Ah, OK, OK.

16:16
Yeah, four understand.

16:17
So there's only for 3G-4G Yeah, yeah, Binet will present the 5G, but we don't have currently on 28 Hall.

16:25
So OK, I'm right understand, Understand.

16:27
OK, consumer, consumer SMF, we don't have this flaw.

16:32
It is really separate flow.

16:33
But Sankar has a question has kinda Sankar, can you please go with your question please?

16:39
Yeah, in the system sub CC is currently deployed.

16:43
I mean is it GCP cloud or in some other environment is deployed and functioning Currently sub CC is on prime, on prime.

16:54
OK, OK.

16:56
And you have mentioned that this matrix is going to handle only SMSMMS under data traffic, right?

17:02
Rating and charging you can which system is like handling voice?

17:06
Voice rating and charging nobility system not not unreal but the Pitt see see I already explained that there are two systems.

17:19
Prepaid voice is handled by online MZ but it is a part of prepaid project.

17:25
Postpaid voice has scandaled has been rated by KB as offline, offline.

17:32
And I explained the voice is not part of this project currently.

17:37
OK, OK, thank you.

17:39
It will be next in the future phases when we also planning to move a voice, postpaid voice.

17:45
But it will not be done in this as a part of this project will be separate project which will run after this project to get rid of of KB rating system and more voice in CCS.

17:57
This is the future.

17:59
It's not current project and OK, just a second somehow you have a questions, please go ahead with your questions.

18:06
Yeah, yeah.

18:07
So the postpaid current architecture provided by Kathy has the GYN 40 interface from PGW to all OLMZ online mediation zone.

18:19
OK, just a second N 40 is coming from SMF.

18:25
Yeah, SMSMF is the not I wouldn't say manual function for that's the we have mission Curry chain 28, right.

18:34
Maybe we have two core networks.

18:36
We have core network for 3G-4G and coordinator for five GSA.

18:42
No, I agree.

18:43
And we have need different element P gateway, it is for 3G-4G for five GSA core network.

18:51
We we are getting charging requests for SMF session managed function.

18:57
I understand that.

18:58
What I'm asking is the GY interface seems to be there already with PGW to OLMZ.

19:05
But I thought you mentioned that we don't have that as well.

19:08
That is also AP.

19:10
No, no, I mentioned mentioned N 28 and 40 NSY.

19:16
OK, fine.

19:18
Not GY, GY we need for prepaid GY is there.

19:22
But what is not there, it is five GSA, it is N40.

19:27
We don't have a tools for N40 and N28 and we also do not have tools for SY throttling.

19:36
OK, OK.

19:38
Because GY we have a lot of stuff.

19:41
Well GY your effort to integrate the GY is the not really small part.

19:46
But there is a notification which we have much more notifications, but I'll turn start talking also.

19:55
OK.

19:56
Yeah, I think as I think N 40 was also shown in the current architecture between SMF and OLMC.

20:04
But from what you what you are saying not online, I'm sorry.

20:07
It is CHF charging function.

20:11
Yeah, I think.

20:12
Let's make it the same because in cases diagram, she kind of cascade 4G5G into one diagram, right.

20:21
So the the SMF and P gateway one for five G1 for 4G, but their their functionality is the most almost equivalent.

20:31
So that's why you may see in 4G and GY or listed, but the 4G is we are we it's GYSY and the file would be is importing in 28, right?

20:42
That's the two interpreted by doing the same thing.

20:48
Yeah, right.

20:50
I understand.

20:50
I mean some of the vendors, like for example, as I know some of the vendors that combine the 4G5G functionality and do it as part of the PCW, then in that case SMF functionality is also integrated into the PCW.

21:03
So that's what we thought probably from Kathy's architecture it could be that way.

21:07
That's what I think Shankar raised the question.

21:09
How are you saying?

21:11
Yeah, yeah, that.

21:13
Yeah, as you see.

21:15
Yeah, that's what I think The SMF, the vendor is different in PGW, the vendor is different in Dallas case.

21:20
OK, OK guys, it is.

21:22
We have a dual bot integration and we are having this, but there is completely separate and we have also separate project to have dual core network.

21:35
And there is about interfaces will stay completely separate for 3G-4G and for 5G understand it all.

21:43
It will be always different interfaces and different actually part of this dual core network which will be used for 3G-4G network and for five GSA network.

21:55
That's clear for sure.

21:57
File G, yeah, in the in the in the case of the file G and I say right, non stand alone, the the SMF will generally use the GY, right.

22:07
But at that time, yes, so so they the non stand alone mode.

22:12
But now it's so when we talk about file G, the data is going to show it's for file G stand alone.

22:20
OK, OK, OK.

22:22
I think when Binay presents the 5G architecture, probably we get more details.

22:27
Yeah, I think we can proceed.

22:28
Yeah, for, yeah, for the for the 5G party we've got yeah, yeah, for for all I'm that is if part of app team is going to do the change and it's kind of a standard solution.

22:42
We, we keep today's Oracle database.

22:45
We, we are in the for the migrator Postgres, we have all this referent data and that's all the logic is.

22:55
Most of the logic is table driven means based on referent data when you generate a separate value and send it to CCS.

23:03
So that's something that we want to mention because this kind of guidelines changing originally is thought of when we introduced CCS, we thought OK, now DRA or DSR so that the GY we do nothing and as soon as the CCS they will handle all the logic, right, we reduce this local functionality.

23:29
But later that changed some of the function being kind of post back or they, they, they said, oh, either CCS cannot do it or CCS it's not efficient for CCS to do the mediation.

23:46
We call it mediation logic.

23:48
So first, for the postpaid CCS product, we still need to do that this table look up, but that's makes developer easier because we are doing this kind of logic today for sub CC.

24:01
If that if we need to remove that logic, then we, we keep it for sub CC.

24:07
But for GY to CCS, we don't do this enhancement.

24:12
But if CCS can can also not able to do it, then we we need to keep the evasion logic before.

24:19
So this one, this part is not a standard.

24:23
It's not the same GY traffic coming from P getaway because we need, we call it Taylor's customer AVP because CCS they sell some of the local we don't we we will apply the charging, but we don't want to do a look up to gather the information.

24:40
So we still need to do a look up in our Amazon.

24:44
And for all these, we have a PPI table.

24:49
I'm not sure if you see the similar.

24:50
This PPI means a performance DV.

24:54
So we track the design, track all the interface, no matter in South spawn and N spawn.

25:01
We track the latency plus the the response code right.

25:06
So in the PPI table we can easily tell which interface is an issue, which system is a facing issue.

25:13
Either they are not responding or the the the error coding the the error response increase.

25:21
We can based on this TPI table to trigger alarm and trigger all kinds of operation related things.

25:32
So for for the new hostpaid Migra CCS migration is a new interface.

25:38
We will need a on this API table so that we can do this same kind of monitoring for this new interface.

25:50
This is the one thing.

25:51
So basically there's a S spawn is PDW, SMSC, MMSC and N spawn is CCS.

25:59
So we need to do the look up, we need to do the on Amazon, need to do the whitelist means that white list means decided when, when the cutover happens for the some number you need to send it to CCS and some number you need to send it to sub CC like gradually cut over to CCS.

26:18
So that's the functionality.

26:23
It's not a it's it's it's mainly needed for cut over and the UAT testing.

26:30
There's another interface.

26:31
It's for for the unblock, there is a question.

26:37
Hi, yeah, thank you Liliana for this OCS config and reference data which you explained right now.

26:45
So I understand this is plays a crucial role in for look up and then for the load balancer to route the data for respective instance of the CCS or SAP CC etcetera.

26:59
This is what I understand.

27:01
So who provisions this OCS config and reference data, how means what criteria you use here?

27:08
I mean for the for, yeah, for the whitelist load of balancing that's maintained by the operation team, right.

27:17
So let's say yeah that we need for the visit number, IoT number to CCS.

27:22
We don't want to send it to sub CC.

27:24
That operation team can can go to the Oracle DB.

27:27
Today's Oracle tomorrow is post GRAS.

27:29
After saying that that's another initiative we are doing.

27:33
I added the number to a table.

27:36
So online we automatically refresh right in 5 minutes.

27:40
So we will see all this number.

27:42
You want to see that the CCS only and that all the checks from from that number will go to CCS and stop seeing it.

27:48
To stop CC, you may configure the number to CCS as a primary subsets as secondary.

27:55
And they would say that to boot the system, but only use the CCS response.

28:00
This, this the table that's maintained by the operation team.

28:05
And then I really we need a kind of a, a, a process, we need to add some number to the table.

28:12
Pretty much these operations tasks.

28:16
There's other table I think you've got to ask is for the the what is called a mediational business logical related reference table is a snapshot.

28:28
This table is used you already be populated by that's much of the owner of the table.

28:35
Like for example, there's a some people pretty stable, like for example, we we need a map MCFCCMNC right to a country code like 302220 to Canada or for one zero or something to Verizon.

28:52
Those table is pretty stable, but they are also whenever there's a new roaming carrier introduced, they will add a new entry that's owned by the, I think key.

29:04
I thought it was the team, but it's like the old carrier relation department, right?

29:11
They they maintain those table.

29:12
They are the owner, But there there's also other table for like a for the sale sale the the, the WH, the HSI part that we already finished which, which cell is allowed of all these smart Harbor that's but whoever the team that owned that they, they need to populate, right?

29:32
We, we accepted this operation table that we need to modify or delayed change for this mediation related table.

29:43
They there's different business owner, so we just use a copy and the copies we are doing you already the daily refresh, but it's configurable.

29:54
You can refresh every a few hours or every every day that and once the the the table is snapshotted to these DB online that we cache it in memory and there's another refresh.

30:12
We can decide the whole like for example the the wireless table we refresh every 5 minutes means if you insert a number in digit DB 5 minutes, 5 minutes later we in the memory online that we starting guided a number to which charging site.

30:33
OK, understand, I understand.

30:36
So, so basically what I understand from this part is 2 part, one part is the this one refers to the master, the migration control, the migration Subs control table to update the whitelist cut over to see to which the route the traffic to CCS or SAP to see the second part which you mentioned based on the MSCM&C routing.

30:58
So these are sorry to interrupt the MCCMACI just want to give a example that that table is stable and it's not owned by the operation team.

31:10
We use SNAP.

31:10
Ah, OK, OK.

31:13
We don't have MCMAC routing.

31:15
Yeah, let me explain.

31:16
We have lots of reference data in our data base and those reference data have been used for different purposes.

31:26
Some of them have been used to build and reach parameters which you are sending which we will send to CCS.

31:34
For rating we have and also database has also this configuration for side by side cutover phases routing.

31:43
We that we have very rich database and probably if you're planning to work on with this system, you need to understand actually our sector of our databases also.

31:58
OK.

31:59
The main purpose I was asking Lydiana is I just wanted to check for for our postpaid project other than this whitelist cutover, will there be any other we have to do to update these for the postpaid service to work only from that point of view?

32:14
Now what what I understand is I think we don't have to do, right.

32:18
I mean just accept the updating the migration service control table, we don't have to do provisioning or update this.

32:27
You don't need to do it is already established.

32:29
Accept this side by side tables in those routing depending which team will actually lead side by side and everything.

32:36
If you're leading side by side and cut over spaces then you'd only this table for routing has to be updated.

32:44
All other reference tables it have already built process for updates.

32:50
You don't need to worry about other, but you need to to actually do some queries and look up to be able to build those enriched parameters which CCS system requires for rating.

33:08
Understand, understand.

33:09
OK, it's clear.

33:10
Thank you, Lilia.

33:11
Just one question.

33:12
Is this mediation zone online Mediation zone, is it a CORDS product or it's a home grown application?

33:20
Is it from another company?

33:21
It's very perfect.

33:23
Not a home grown.

33:27
Yeah, it's the company called Digital.

33:29
They did, they did the road Digital Rd.

33:31
mediation zone.

33:32
OK, Yes, yeah.

33:34
OK.

33:34
And is this the same application or different instance is used in offline mediation or it is different?

33:40
It's the same winner, the CTF offline mediation with a few other installations all used.

33:47
OK, So different version, it could be different version, but from the same winner, same one different, Yeah, OK, OK.

33:56
And it's on Prem right Elton for both of them now it's on prime.

34:02
I think of offline they tried the the the Kubernetes version, but I'm not sure what's the latest.

34:09
They have offline team.

34:11
OK, OK, oh, see they they already share campus session.

34:15
They are moving to camp.

34:16
We are also going to move to cloud.

34:18
We are also planning to get rid of digital route and build everything on the cloud offline mediation.

34:27
I, I saw this morning they share all those transfer for camp.

34:32
They're using camp framework to actually move everything from Prem to the cloud.

34:38
And also our team is working in parallel with this project, we are working on other initiatives to move everything from digital route to in house build Blacker loud.

34:50
OK, OK, thanks.

34:52
Alia.

34:52
At least for the timeline when we do a postpaid delivery, the side by side.

34:58
So it'll still be a a digital route, right Or will there be any cut off time to move on And it will not be digital route.

35:07
It will goal is to get completely out read of digital route.

35:13
OK, but during but it's not going to happen during this project.

35:17
You don't need to worry for you.

35:19
You will do everything on digital route later we will do migration.

35:25
Perfect.

35:26
Yeah, thanks that helps any any other question.

35:45
So for it's I think I called a pretty much the the 3G-4G interface and this is regarding the how the CCS color were.

35:54
But for when we come to SMS, of course here is the SMSC MMSC.

36:00
There's also DSR when it comes to there's no SY good to use and there's no notification.

36:07
But we still have these OCS configuration for the S mainly for SMS and that we see that to CCS, the SMS one already mentioned for US roaming, we now send it to CCS directly.

36:21
So we need to build a separate interface.

36:23
I think today we already have for WHSC for CCS, we already have a catch, we call it a catch rating interface is a damage or event based charging For these US roaming, I'm thinking we could reuse the same, but this had to be aligned with CCS team, right?

36:42
So that we it's a GYUV interface when you're saying that the US roaming SMS to CCS the bit so that they can charge the US roaming SMSMMS is standard.

36:55
We we got the the RO interface from MMSC, we send it to CCS.

37:03
There's no roaming for MMS hash rating right where we is applied to we we're talking about South morning Springs.

37:14
That's the happy path.

37:15
Everything works.

37:16
So the switch set of the request that they got the charge successfully perfect and nothing to worry about after that.

37:22
But there's a case.

37:24
Let's say the request we send it to CCSCSSCCS time out, right?

37:29
All the worst case online is crashed.

37:33
But we we good news, we do have a dual redundancies never happened.

37:37
I'll just make up these situations that the real time failed.

37:42
So the real time charging field either because of CCS or because online MSA or because of any other reason the PDW, SMSC, MMSC will mark the usage as that set of flag or the usage didn't charge successfully in real time.

38:02
So there's a separate process for PD with the CGN and therefore SMSCSMSMMS is in offline site, they will open the raw Cdr, right switch to general Cdr and they open the raw CD and take all these SMS didn't charge us.

38:21
That's the same thing happens to the US roaming because US roaming we are, we proposedly when we see, we detect the US roaming, we actually tell the SMS the OK, we cannot charge it.

38:32
So they will mark as a non rate not not rated.

38:36
So for, for the, for this catch of rating and offline that we need to extract this usage, they will send it back to online.

38:47
So we need to send it to CCS, right?

38:49
So make sure that the the revenue is is not lost.

38:52
So it's a big kind of a delay the charge.

38:55
It's more like the old offline charging time that we used to switch role records to do charging, but it is only applied to the the record that filled in the real time charging.

39:08
So for CCS online that CCS we do have even the base charging today, either we we use the same interface or we introduce some REST interface that's to BD.

39:23
But I most likely I think we we may use the the event based charging.

39:28
Just saying that all the category can records.

39:31
So that's a that's a second type of usage about related to the post PDP gateway SMS and MMS, I'm not sure.

39:43
Oh, OK.

39:44
I'm I'm expecting questions here.

39:46
OK, go ahead.

39:48
Yeah, Elton, one question about that offline charging, right, which is coming as a tap in or raw Cdr to CCS.

39:55
So do you do any deduplication or correlation as such to verify whether it is a?

40:00
Only unique Cdr center CCS, the duplex.

40:05
I think offline they do some duple check that the when they receive the raw Cdr, when this they check, they open the raw Cdr.

40:13
Let's say we take the SMS example, they grab the SMS raw reference from SMSCI know they have some duple check.

40:21
But I don't know if the duple check happens before they extract the usage or after maybe 99% of the usage or really really can try to success right?

40:32
Unless there's a production outage, we have it to do it.

40:35
We call it a bypass means online that we would not talk to sub CC.

40:40
We would just blindly grow up around the quota.

40:44
So then offline Saturday we will see all the raw SMS in the real time charging, right.

40:49
But if when the all the system is up running, most of the SMS is already trying to successfully they only seen a few right?

40:57
For whatever reason it could have because time out or something the dual check if it we don't we don't have.

41:05
That's a good question.

41:06
When we receive the type of rating records, we will, we will not pick up the same batch.

41:11
If this give us a 2 batch with the same name, we would, we will not pick up.

41:16
But if there's duplicate check records within two batch, we will still send it to CCS.

41:24
There's no Dupree check for online answer.

41:28
OK.

41:28
Yeah.

41:29
And is there a scenario for inbound roamer which is handled by online mediation, the inbound roamer, you mean non Telus customer, right.

41:39
Non Telus customer, Yeah, that no, I don't.

41:42
That's the roaming, the roaming settlement, that's Infinity.

41:47
OK, OK, that's fine.

41:48
And one last question, you talked about US roaming.

41:51
So is there a scenario for rest of the world roaming like Europe or something?

41:55
Or it's not at all.

41:56
That's all out, that's standard.

41:59
Oh, US roaming is only for US roaming SMS.

42:03
We can try the real time, but there there's a this is a long time.

42:11
There's a long time requirement.

42:13
This lets you at for SMSC, you cannot detect AC US border.

42:19
OK, so that's why.

42:22
But they still want to avoid US roaming charge for being like a border.

42:29
So that's why we use Happy.

42:31
OK, excellent.

42:34
Elton, regarding this SMS event charging, you mentioned understand that for like US roaming and all about the domestic SMSS that are sent inside Canada.

42:47
So those are real time which interface is used?

42:50
Yeah, it's which the our interface the the SMSC.

42:59
Oh you mean for SMSC?

43:00
Also the RO interface is used to send a normally cap map or some other interfaces are used for SMSC for SMSS, right?

43:08
I mean it's RO is a real time is it diameter interface?

43:12
What is the event based right CCR event.

43:16
OK, OK, OK.

43:19
Yeah.

43:19
So for for for in case for testing purposes to test SMSS then what we have to consider is only the auto interface, right For testing you have to cover both the RO and plus catch rating right to make sure CCS can can handle the catch up rating traffic.

43:42
So RO is happy pass, yeah.

43:45
But I see tools for SMSMMS and GY, because we needed this for prepaid interface.

43:52
We use matrix test framework and we actually have simulation from networks to CCS.

44:00
We have test cases, but our test cases actually only for prepaid.

44:05
You actually need to go to use the same framework and other test scenarios relevant for the postpaid.

44:13
But we already started building the suite and we have a lot of already done stuff for prepaid.

44:21
This is only actually told which we currently have been using for prepaid.

44:26
We have used this for we also have for multi for prepaid.

44:30
But you don't have voice functionality as a part of this project.

44:35
But you can use for MMSSMS and data, you can use Matrix framework for 3G-4G data, you can use Matrix test framework and whatever we build for prepaid.

44:51
You just need to add additional test cases for postpaid.

44:55
But this is not case we don't have for N40, we don't have for SY and N 28.

45:02
We don't have this capability at the moment.

45:06
I understand.

45:06
OK, perfect.

45:07
This answers my question.

45:09
Then also it'll be great and it'll be helpful if there's any document explaining about this matrix test platform tools and how it works.

45:20
It'll be really helpful for us to under for our test team also to understand this.

45:26
After this call, if you have any documents, kindly share and we'll read through.

45:31
And CCS team has documents.

45:34
CCS team will share with you guys.

45:36
OK.

45:37
OK.

45:37
Thank you.

45:38
OK, so I mean, who is the contact?

45:40
Do we from the CCS from Tellus side?

45:43
I don't know.

45:44
You need to confirm with Saki.

45:45
I really don't know.

45:46
OK, then I will contact Saurav.

45:47
Yeah, I will check with Saurav.

45:49
Yeah, no problem.

45:50
Thank you.

45:50
Lilian.

45:50
Saki.

45:51
Saki, you need to check with Saki.

45:53
Yeah, OK.

45:54
I mean from our side, Saurabh is the point of contacting any contact Saki or internal.

45:59
That's why I mentioned Saurabh from OK, OK, yeah, definitely.

46:04
We go through the saddle.

46:05
That's why, Yeah.

46:08
And and Elton, how long you need because we need to give a little bit time for 5G and say no, That's why.

46:13
That's why I try try to ask.

46:15
I think we still have 14 minutes.

46:19
Is that the time right?

46:21
So I can, I think I covered the overall diagram and I mentioned the the interface that postpaid need to talk to and also the it's the real time plus catchable rating we need to handle, right?

46:35
I mean catchable rating maybe only count like 10% or utilize 5% over the traffic or I think in the in case so that no outage is 99% is really written in real time.

46:48
But when it comes to development effort that make that we still need to develop the the what is that rating interface?

47:01
Yeah.

47:01
So, so let him talk about the 5G quickly and we can always do more session like more detailed session when it comes to the the certain interface.

47:17
But it is I hope I can do give you guys like an overview of the post page for mainly the 3G-4G architect.

47:29
Thank you very much Elton.

47:31
I will send you the link for folder in which you can put the documentation which you share from this meeting.

47:38
Vinay, are you still on the meeting?

47:41
Yeah, OK.

47:42
Can you please share your document and provide high level design for five GSA which we actually currently have in currently in the production we have only Telos 5 GSA customers.

48:03
Yeah, I share my screen, we cannot see anything.

48:07
Vinay, can you see right now?

48:14
Yeah, it's perfect now.

48:15
Thank you.

48:17
OK, so I'll start with the five GSA architecture.

48:24
So to give you a brief, so right now we have 4G and 5G, but the standard flow of the you can say the the data and the and the workflows logically we kept it the same.

48:40
OK.

48:41
So to start with the network architecture, I'll explain you what all components we are actively interfacing with.

48:48
So we are actively interfacing with PCF, we are interfacing with SMF, we are interfacing with CHF and RF right now in five GSA we don't have, but there were discussions, but as of now it's not there, OK.

49:06
And for us these three are the main interfacing applications for our application.

49:15
And this is a very standard network architecture we have.

49:20
So moving forward I'll go to the our application details.

49:23
So before like 5G we were having this kind of network architecture where we'll, we were dealing with the camel and diameter 3G-4G part and the CDRS were created by B gateway.

49:39
And now in A5 GS architecture, we are having N40 interface that is directly interacting with SMF and through the NCHF and 40 STPI and CHF.

49:54
This this module which we are seeing which is responsible for getting the session request and then processing and then sending it to the SAP CC.

50:06
And then once the usage is related, then it sends over to the RAW and then reduces to the billing domain.

50:14
OK.

50:15
And moreover, if we see the art on a very much high level architecture, so it's a pretty much the same.

50:24
So this is a kind of architecture that will be in in future.

50:29
So right now we have SAP CC in place, but in future we, we are planning with the CCS.

50:37
OK, So SMF is going to greatly interact with the postpaid and 40 mediation layer over N40 and we'll be using the same routing mechanism which we'll be using for 3G-4G as Eldon explained, OK.

50:55
And this PCF piece will be completely new because this is not existing right now.

51:02
So right now we have contraction with PCF over SAP CC that is still we are using the whole you can say the throttling part and policy control mechanism part over 4G workflows.

51:17
OK, so I'll explain you a bit maybe with this diagram.

51:27
Can you see this diagram?

51:30
OK.

51:32
So this is a complete medication layer we have right now, OK.

51:38
You can see the SMF is there.

51:40
NRF is not functional right now.

51:43
They are not integrated with SM, sorry, NRF, we have great connection with SMF to CHF workflows and there's a common component you can see, right?

51:52
So common component means that for the notification, OK and for all those throttling and policy control related stuff, we are still using the common 4G kind of workflows and that that is already there.

52:11
So what we do is whenever we get a request, we establish a session with SAP CC and whenever we see that some notifications or any policy related thing needs to be triggered, this 5 GCHF triggers these common workflow that still resides, you know, 4G-4G servers.

52:33
And because since that is a common for I can say the all 4G and 5G functionalities, so we reuse those components and then we used to send across all the standard.

52:45
Yeah.

52:48
Hi Bennett, there's a question.

52:50
So the CHF component which you're showing, is it part of the metrics this year?

52:55
So or is it part of it's independent entity?

52:59
It's independent same like 3G-4G like but it is it part of the OLMZ to mine or it's not?

53:06
So the platform additions only the same online MZ we mean for 4G5, for 3G-4G and when we call CHF that is meant for 5G.

53:18
OK, OK, OK, I understand.

53:20
So basically it's an independent entity from the lists.

53:24
Yeah, servers are completely independence, even workflows and everything is completely independent.

53:32
OK, understand.

53:33
Thank you.

53:35
So, so on a core basis, the core functionalities are being handled by the 5G workflows, but still the common functionalities we still routed back to the common workflows using the diameter, OK.

53:51
Still you can see a lot of integrations with the SAP CC, OK.

53:57
And going forward, we'll be planning to put instead of like CCS and the same as the I can explain, right.

54:05
So the same whitelisting logic will be implemented in 5G as well.

54:11
So if you if you look from a logical perspective, SO4G and 5G logically is almost the same.

54:18
OK, Only the common components we reuse is a 4G common workflows that that is mostly for policy related, maybe throttling, unblock, unblock.

54:28
So these are the very IOU.

54:32
Any questions anyone, uh, in the, in the architecture data, you have mentioned this reference DB right from BSS to mediations.

54:44
And what is that?

54:44
Is that the control, uh, table or something else?

54:47
Reference DB.

54:48
So reference DB, uh, is for, uh, various table lookups like, uh, cell IDs.

54:55
And also there are lot of table related to.

54:58
We have a list of them, the same tables we have in 4G5G As for sorry, 3G-4G as well.

55:03
So tables are the same OK, we use the same log.

55:08
Whenever we get a session request we have to enhance or enrich few of the parameters.

55:14
So in order to enrich those parameters sometimes we need few reference lookups OK plus for outlook whitelisting and some small kind of informations like we have few functionalities like KPI.

55:30
So all these reference tables we keep as a common and 3G-4G workflows and 5G workflow refers to the same.

55:39
So we don't keep it the sort of table are the same database is the same.

55:44
We don't have our separate databases for 4G and 5G, so it's a common.

55:49
So whatever Elton, I explain we the same schema and everything we'll use for the 5G as well.

55:57
OK, No, I just sent one recent question to right now what you explained.

56:08
So that means that for example for 5G scenarios today, in case if you have to, I'm not sure for the side by side or the cutover for whitelist cutover.

56:21
So therefore also will refer to the same reference tables from whatever Elton explained, right.

56:26
Yeah, as of now, it's it's the same table.

56:29
It's up to up to the team how they want to completely go through the cutwog strategy.

56:35
But as of now, the table and the whole database is the same.

56:40
It can be reused.

56:42
Yeah.

56:43
So basically the theory we'll reuse this, try to reuse this and then decide where to move the traffic.

56:48
Yeah, yeah, that can be done.

56:50
I see.

56:50
I think that Cataly will first explain migration strategy and how it will be used.

56:55
It will be used little bit more complex than we use in the past because the all migration strategy is very different.

57:05
To really go, we first need to understand what is requirements and after that we will see actually what has to be modify or added in this table.

57:16
But let's first have a meeting with Kathy Lee about migration strategy.

57:20
And after that next step we need to understand what has has be need to enhance here on our side or not.

57:30
I understand, Liliana.

57:31
OK, thank you.

57:33
And, and for the so the is, is that the CHF which is in mediation zone is a separate component.

57:41
It's not part of digital route, right.

57:44
So this is a separate everything is digital route.

57:47
Everything is digital route.

57:49
OK.

57:50
So CHF is also part of the digital route.

57:56
Yeah, currently everything, currently everything what we're talking on today meeting is based on digital platform.

58:02
We are planning to really get rid of Digital Route, but it is a project in progress and this project has to be finished before we actually migrate this on the new platform completely independent of Digital Route.

58:19
That's fine.

58:19
I, I, I thought CHF is a separate component which is not part of digital route.

58:23
Yeah, I understand that now.

58:31
So for the session requests, whatever we are getting from SMF, we'll be processing those to the CCS or SAP CC.

58:41
In case of failures we have default handlings in our in our CHF workflows and also for we have a separate flow where SMF used to push those files via mediator VM and we used to offline process those files and send it for a catch up.

59:01
Actually it's not exactly the word catch up.

59:03
So they use SMF buffered files.

59:05
So whenever SMF sees that there is some connectivity issue or CHF is not able to respond, then they start buffering those usages as per the standard 3 GPP.

59:17
And then they used to push those files to some mediator PMS we have in place.

59:23
And we used to and they used to push those files to our offline servers.

59:29
And from there, we used to pick those check for like say like file level duplicate check.

59:34
And then we used to process and send it to to the SAP CC right now one more question to go to Benai and Liliana.

59:48
So now that product we'll have to develop the tools for testing for the 5G, for example, for N40 and N28.

59:56
So if we follow the standard NCHF interface based on TCPPI, think that should be fine, right?

1:00:02
Because what I see here is.

1:00:04
You have something called 5G workflows which CHF is translating.

1:00:08
So I'm not sure how this 5G workflows for the 5G from CHF to the further CCS, what interface is used whether it is again a custom based like what or Telus developed based 5G interfaces or not.

1:00:23
So this is Vinay, we'll explain this.

1:00:25
We have enriched parameters.

1:00:27
It is not completely standard.

1:00:29
We have some enriched parameters, but then you OK, All right, So OK, I understand.

1:00:34
I understand.

1:00:35
OK, then I'll wait for the nice explanation.

1:00:37
Thank you.

1:00:38
So, so for now we have the all the protocols we have implemented for this 5G is CGPP standard.

1:00:48
OK.

1:00:49
When we talk about the enrichment, yes, we will have to add new parameters as per the need.

1:00:55
Yeah, please somebody has a question.

1:01:01
Yes, SO5G usually compromise like in a low, ultra low latency, right.

1:01:08
So here I can see control plane and user plane, we have a separation.

1:01:14
So how it's like an impact the latency performance in.

1:01:21
So I'll first complete the previous question and maybe then I'll come to this.

1:01:25
So with respect to the protocols, we are very much following the standard protocol and when we need enhancements, obviously we will have to enrich those parameters in order to in getting In Sync with the SAP CC or CCS.

1:01:44
So the enrichment part is that we have to add a custom kind of parameters into the standard YAML files.

1:01:53
OK.

1:01:54
So as of now, the whole architecture is almost we are following the standard interface.

1:02:00
So even if you are building some testing tools or something, if you follow a standard HTTP two protocols, that should be good.

1:02:09
You should be able to test everything.

1:02:11
And yeah.

1:02:13
And yeah, so just to sorry, I mean, OK, so understand, I mean, we also would like to understand, you know, what are the additional elements?

1:02:27
So the information elements that we have may have to pass for postpaid specifically if you can help us identify this and, or if there's any documents.

1:02:35
Yeah, yeah, it's already mentioned there in the document.

1:02:38
Please share it.

1:02:39
Just.

1:02:39
Yeah, sure, sure, sure.

1:02:41
It, it will be there.

1:02:41
Yeah, everything is there.

1:02:43
OK, understand.

1:02:44
So then I think it'll be useful for us.

1:02:48
I I think the next question was regarding the latency.

1:02:51
OK, so right now we are in a control plane and there are timers already set over the SMF and we also have our timers, OK and so mostly we used to get a response in very quickly maybe in milliseconds like 20-30.

1:03:09
Within 100 millisecond we used to get a response.

1:03:13
But if in case we see any delays or something we have process in place like so there are few default configurations at SMF side plus we also have some default configurations at CHF side.

1:03:26
Few of the scenarios are where the the CHF servers are overloaded then we used to send a quick response to them that we are overloaded.

1:03:34
Please try some other instances with error code 429.

1:03:38
So there are mechanisms in place which will anyway avoid any kind of further delays unless there is some major hiccups in the system like CCS or SAP CC itself is a down CC and we are in a some maintenance window, right.

1:03:55
So apart from that we, we have all those scenarios included which will avoid the large latencies in the network control plane.

1:04:04
And Vinay, we cannot hear that if something is completely down in this case SMF will create.

1:04:11
Yeah, they are, yeah, yeah.

1:04:15
And we will replay this usage after that we have catch up flow for five GSA and we'll also have SMS 5 usage replay priorities to both both processes are yes and we yeah, there is of all those documents in there are both processes have been explained in this document.

1:04:38
OK.

1:04:40
Do we have like any mechanism like an in place to ensure like an high availability of the networks?

1:04:46
If, if how the system handle the failure of a network function, any specific mechanism?

1:04:54
So if we talk about failure, right?

1:04:56
So we'll have to 1st identify, there are three layers as of now.

1:04:59
First is SMF layer, another is a mediation layer.

1:05:03
And then we have a ACHF code layer that is CCS or SAP CC, right?

1:05:08
So every at every level, right?

1:05:10
For if we talk about the failure at our rating level, we have set of mechanisms like we, uh, give default quota and we also have a different retry mechanisms, OK, in place, uh, if we talk about the mediation layer failure itself, again, SMF has some default handlings.

1:05:29
So, uh, the if we talk about the availability part, right?

1:05:34
So right now we have 2 sites, code and ball in production and we are expanding 2 new sites that will be mutual and fit fit.

1:05:41
So from SM FS perspective they have a kind of failure handling mechanism that if they first try on some any hood instance and if if they return with some error on they are not able to respond it well then they will try to the other Geo redundant site.

1:05:59
So these things are already in place.

1:06:01
OK.

1:06:03
Are you using any codes specific tools to validate these scenarios?

1:06:09
I think there is no such tool to test this, but we have to kind of plan these things that if you want to test Geo redundancy, right, then you have to maybe down your view of the instances wherever you want to test, OK.

1:06:28
And in in testing involvement, I think SMF mostly controls those redundant Geo redundancy kind of stuff.

1:06:38
So we, we act as a server, OK.

1:06:40
So we don't have to much care about the redundancy part.

1:06:44
We have to just be available.

1:06:46
So SMF used to test those.

1:06:49
So we have to just support them that, OK, if you want to test this, then we might have to support with the we'll have to temporary down few of the services and they will test whatever they want to test.

1:07:09
Any other question?

1:07:20
One question which you mentioned like in case for a bypass scenario, like if the CHF is completely down, what you explained is that SMF creates the CDRS offline.

1:07:32
How are these transferred further to for consideration into rating or to reconciliation?

1:07:41
Yeah.

1:07:41
So for when CHF is completely down.

1:07:44
OK.

1:07:45
Yeah, there are two, two steps or two kind of methods in place right now.

1:07:52
First one, if the CHF mediation layer and the teller CHF is down and we made it down because of some maintenance window.

1:08:03
So we have option of a bypass workflows, we use those workflow to give some default quota, OK.

1:08:10
And then further we can use that as a catch up, OK.

1:08:13
Another one is that when CHF, Telus CHF is down without any kind of planned activity, right?

1:08:23
If if it's down due to some technical reasons, then in that case SMF has a way that they and that is a very much standard 3 GPP practice that they will start buffering those usages and wait for some time for Telus CHF to up again.

1:08:40
And if in some limited duration they comes up again, they will replay those files over standard and 40 interface.

1:08:47
Otherwise, if they reach some threshold, then they're going to create push all those usages into a file and those files will be pushed to a.

1:08:56
There is a mediator VM.

1:08:57
So this mediator VM is in between SMF and Tela CHF.

1:09:01
So this is a kind of mediator box where they used to push those files and again from there they used to push those files to the CHF servers.

1:09:11
From there we used to pick those files, create a Cdr for reconciliation and parallely we also send it for a buffered file catch up, OK.

1:09:22
And then we used to again send it to the SAP CC for rating and and then SAP CC creates a TDR and then so the whole process of reconciliation is done from raw versus rated.

1:09:35
OK, I understand the the reason I asked is in the architecture flow, there was a flow from to the CGF component.

1:09:44
Then for CGF component, it was going to the cloud camp for the offline the raw Cdr.

1:09:50
So I just wanted to confirm that this part will not be followed.

1:09:54
But for the bypass scenarios, what I understand now is for all the bypass scenarios as well, you'll still go through the CHF and the standard and CHF interfaces used to push these records also to either to CCS or to SAP CC for further rating.

1:10:08
So for 5G world Telus CHF now is responsible for creating CDs.

1:10:14
So, so for any flow, whether it's a normal, we can say standard and 40 request or the request is coming by a file based, preferred file based kind of approach.

1:10:26
For both of the cases, Telus CHF is going to create those CDRS, raw CDRS and the CGF is going to pick those and process the way they want.

1:10:36
OK, so we create those files and that is being further processed.

1:10:41
And for the rated part, CCS or SAP CC create those and they again get processed by different interfacing systems.

1:10:50
OK, OK, I understand.

1:10:51
So basically then then the rated CGS will still be generated by them and then pushed further to the offline mediation for other purposes?

1:10:59
Yeah, OK, understand.

1:11:01
Thank you.

1:11:08
So for 5G part because of all the policy mechanisms will still being handled by the 4G common workflows.

1:11:18
So that is kind of reusability we are doing right now in terms of reference tables, databases plus the common workflows, 4G5G workflows as you can see here, right.

1:11:31
So this is something we are using a common and the rest of the control plane messages we are greatly adding to our system.

1:11:38
Yeah, please, yeah, please go ahead.

1:11:47
No, it's just Benjamin is leaving the meeting.

1:11:50
OK, No, no question.

1:11:52
I cannot see like what messages and all that's OK.

1:11:55
So if anyone has any question because, uh, uh, from a very, uh, overall architecture wise, uh, I explain whatever, uh, needs to be from a very much a broad perspective.

1:12:09
If anyone has any question or anything, still we have 15 minutes, right?

1:12:14
So sure, yeah, team, if you have any questions, let's let's ask.

1:12:19
And Vinay, this documentation what Nikhil and and myself, you know, we working, we created, I think Nikhil did a great job.

1:12:30
We created that product folder.

1:12:32
So if you can, don't worry about that I, we will start without the folder.

1:12:38
We will start all documentation after this meeting in the folder.

1:12:42
Don't worry about it.

1:12:42
Thank you so much.

1:12:43
Thank you so much.

1:12:46
Any final questions the team has?

1:12:56
All right, I think you I at least we could now for the we have the basic understanding about how things work sort of as we receive the documentation and then we read through details.

1:13:10
Maybe probably we will have more questions then we may request for another session for clarification.

1:13:17
Yeah, that's fine.

1:13:18
That's, that's fine.

1:13:18
Mahesh.

1:13:19
We can update the update the week to week 3 plan based on it.

1:13:23
Yeah.

1:13:24
OK.

1:13:24
Yeah, sounds good.

1:13:25
It will be good if you send questions in advance to us by e-mail that we actually now on what what we need to focus to answer the questions.

1:13:38
Absolutely.

1:13:38
So we'll do that for next week.

1:13:41
I think with the follow up session, we will send the questions in advance.

1:13:44
Yeah, a couple of days in advance that actually we can prepare and provide answers in the meeting.

1:13:51
Sure.

1:13:51
Liliana, thank you.

1:13:53
Yeah.

1:13:54
But this was a great session.

1:13:56
I appreciate to you spending time with us, Binet, Liliana and team.

1:13:59
Thank you so much and we'll connect very soon.

1:14:04
Have a great have a great weekend.

1:14:07
Thank you.

1:14:07
You too.

1:14:08
Thanks everyone.

1:14:09
Thank you.

1:14:09
Thank you everybody.

1:14:10
Have a great weekend.

1:14:11
Bye, bye, bye.

1:14:13
Goodbye.

1:14:14
Thanks.

1:14:15
Thanks everyone.