Clarence Reynolds, Contributing Editor, TelecomTV (00:06):
Telecom APIs play a crucial role in ensuring that mobile applications run seamlessly. Here to explain more is Pedro Diaz Telco, pre-sales consultant and Ruben's AIO solutions architect, both at HPE. Hello. Thanks for being with me. And Pedro, I want to start with you because here at Mobile World Congress, APIs and open gateways are a hot topic. What problems are these designed to address?

Pedro J. Diaz, HPE (00:34):
Well, the main problem here is the monetization. The fact that you have API that does not mean that you will be able to monetize it. Okay? So this is one of the key problems that we are facing with our approach to the monetization of the telco camera APIs using open gateway.

Clarence Reynolds, Contributing Editor, TelecomTV (00:51):
And so how is HPE helping Telcos monetize?

Pedro J. Diaz, HPE (00:55):
Let me show you in this presentation. Okay. Telco APIs and the Open Gateway Initiative are proclaiming to have a very simple APIs with one line of code and share between all the tel operators. So the APIs will be coming to all the tech operators. So in order to monetize them, you have five options. The telcos has five options direct to the customer exposing these APIs directly to the customer to create an integration platform to code along using verticalized specialists like healthcare, like industry and so on. Or even to integrate the marketplace on the hyperscalers and so on. In any way, they will face the same challenges. The API is very simple, but the network of the operators is not so simple. So in order to do that, you'd have to simplify the network, be able to use the cloud technologies that the hyperscalers and the marketplaces are using, and also be very fast in adapting the services to the customer. Because at the end of the day, who will require these APIs are the enterprise customers. And this is key.

Clarence Reynolds, Contributing Editor, TelecomTV (02:07):
So can you show me a practical application for this, A use case?

Pedro J. Diaz, HPE (02:11):
Yes. Okay. So at hp, what we are doing now is to envelope everything in a form of packages. So we create a packages using our technology. The technology is 10 years old, so we are not talking about something brand new. The fact is that we use in the since 10 years, highly abstracted APIs, cloudification cloud native components and so on, is giving us advantages against other vendors in order to implement that. So what we do is we create a package, the package include the call control, the media, control the portal for their term price, and also the different companies that you create and deploy them. Okay? And in the demo that we will show today, we are using mainly one simple use case. Everybody knows, for instance, the Amazon type services. Sometimes you're not at home and maybe Amazon goes to the tele operator and says, look, I have a problem. I want to put in contact the parcel driver with the person who is expecting to receive the packet and make an anonymous call without having the information with that. For that reason, you can use the camera APIs, very simple APIs and so on to drive this concrete use case. We have more use cases, but this is what we can show in a minute. Okay.

Clarence Reynolds, Contributing Editor, TelecomTV (03:42):
As a matter of fact, let's take a look. Rubens is going to show us the

Rubens Aloisio, HPE (03:44):
Demo. Yeah, please let me first explain how the environment works. So our click to dial service is a cloud native application, so you can run it on any Kubernetes flavor. Okay. What we have prepared for this demo is a vanilla Kubernetes running on a private cloud and we have also a Kubernetes that is running on Google Cloud. What we are adding on top of economical and traditional deployment is the GitHub's way. What does it mean that you can declare on a Git repository what you want to deploy on which cluster and then the system will reconcile the resources. Okay, let me show you the environment. Let me do log in. Okay. This is our private cloud and this is our public cloud. Both of them are empty. Now what we will do is taking our descriptors and just a really, really quick overview of what we are doing here. We are using simple amor files where the operator can declare what version of the application should be deployed on our clouds. You can have multiple versions so you can do rolling update and progressively maybe update all your clusters. Now what we will do is a simple

(05:45):
Commit and push on our Git repository and we will see our application appearing here. What is going on under the hood is on commit after our commit. That's a pipeline GitLab pipeline triggered that. It is invoking microservice develop match speed that is preparing the input for flux cd, that it's an open source and widely used tool for the continuous deployment. Look, we have our microservices that started to appear. We have our media function. We have read SDB that it'll be used by the GUI for the tracking. It's a part of our solution. This is the microservice that it is hiding the network complexity, exposing the simple APIs nor bound. And last one will be the GUI That will appear in a while because in this kind of deployments it's also important setting dependencies between our services so you can have a given order during the installation and control what is going on. If we look also at our Google cloud, POS appeared also here. Okay, so we are controlling two different clouds using the same Git repository and the same set of descriptors. Now our portal was not ready because, but now appeared. We will do a login using an admin profile. This is simulating the business that purchased this service. Amazon as a reference like we said before. Okay? So can control what is going on.

(08:00):
We have

(08:04):
Three class of users created. This first two will be used for the first use case, okay? Simulating the guy that was not at home and the driver, these two users are internal users, can be used for testing the platform or for whatever reason, and this is the last user that we will use for the multifactor of integration. Okay? We will see in a while how it works. We have also the tracing so the administrator can control what is going on so it can control all the calls and what is going on on the platform. Okay? Now let's take a mobile phone and this is the same portal, but this time we will log in with the user profile. Okay, here we are simulating Tony Stack that was not at to, and he received this notification and just clicking this button, we start the call, okay, between him and the driver, that should appear here. Arrived. Can you answer please? Yes. Hello? Hello, can you hear me? Yes, I can. Okay, driver. Sorry I was not at home. Can you please come back? We'll be right back. Thank you very much, sir. Alright,

Pedro J. Diaz, HPE (09:49):
So in five minutes we set up a complete business oriented services using very simple camera APIs that it has a real value.

Rubens Aloisio, HPE (09:58):
Okay, now let's go to the second use case. Again, the portal was not there, but now appeared and here we are, simulating guy that have to do a money transfer and he asked for an additional level of security, okay, multifactor authentication. Okay, in this case it will push the button. Please, please answer, answer please.

Clarence Reynolds, Contributing Editor, TelecomTV (10:39):
1, 3, 5, 7, 9, 0.

Rubens Aloisio, HPE (10:44):
Okay. And click here, we can verify if the code is correct. Okay, now you can check your bank account and see if money arrived. Excellent.

Clarence Reynolds, Contributing Editor, TelecomTV (10:54):
Excellent. Thank you both for the demo.

Rubens Aloisio, HPE (10:56):
Thank you very much, sir. Pleasure. It was a pleasure and I hope you enjoyed.