Tony Poulos, TelecomTV (00:07):
Hi, Tony Poulos here at TelecomTV's AI Native Telco Forum in Düsseldorf. And with me today I have Nilesh Parikh, who is the SVP of AI and Strategic Growth at Mavenir. Welcome, Nilesh. And also Shawn Zhang, who is the CTO and co-founder at Sanas. What a great combination. I've just sat and watched the presentation that you just did and quite in awe. But Shawn, let me start with you. Why is voice AI becoming such an important opportunity for the telecom operators right now?

Shawn Zhang, Sanas (00:38):
Great question, Tony. And I think first off, we're looking at a few numbers. Right now we've got 42% of the population preferring to call over WhatsApp, actually through their mobile. We've actually got 89% of the population complaining that they have call quality issues on their network. And the final number to close this all is that there's been a $25 billion voice ARPU decrease year over year. So right now, telcos are at a losing battle, but with the rise of AI, we can see that voice becomes not just the connectivity layer, but also the software layer where we're able to provide extra value to our subscribers.

Tony Poulos, TelecomTV (01:21):
And Nilesh, what does the Mavenir-Sanas partnership that we just announced bring to the market? What's the value?

Nilesh Parikh, Mavenir (01:28):
The value is really what Shawn is saying. The space technology is now matured, voice is declining, voice revenue is declining, and we want to change that. And the market is ready and AI is available now. And we are trying to insert AI into telcos so we can shift that paradigm for voice revenues. And using AI into voice path, we are now able to offer very innovative solutions, the solutions that Sanas is offering, into the telco networks for the consumer, for the B2B and the government markets.

Tony Poulos, TelecomTV (02:07):
So Shawn, why is the mobile core playing such a critical role in delivering voice AI natively, for example?

Shawn Zhang, Sanas (02:14):
So I think this is where truly being an AI native telco will give you that structural advantage. A couple of things here. So one, as we think about telco, you really want to care about total sovereignty. You don't want data to ever be leaving your network. You don't want your data to be exposed, being used for training, being used by other AI companies. You want to containerise that within your network. So you want to keep privacy as a number one priority, but two is also going to be on latency. You're never going to say, "Hey, can I have more latency in this call?" Well, you can have minimal latency when that extra AI capability is built in directly into your network. So you won't have to worry about the third party cloud API leaving the network, where your voice has to travel from one geography to another geography to a data centre to be processed and then back.

Shawn Zhang, Sanas (03:03):
When you do it directly in that call route, you're going to have minimal friction, you're going to have a seamless experience to provide that impact to your subscribers.

Tony Poulos, TelecomTV (03:11):
It seems like a trade-off, because we've been talking for years about offloading everything to the cloud and now we're talking about the software issues in particular, with data coming back to on-prem. So is that really critical at this stage with AI?

Shawn Zhang, Sanas (03:23):
Yeah, I think so, because a lot of it is about the same investments that we've done to build out clouds — how do we build that same infrastructure and muscle directly into the telco? This is how we make telcos back into being hyperscalers, be a rapidly growing industry and making these capabilities just be part of that overall experience.

Tony Poulos, TelecomTV (03:42):
Well, Nilesh, looking ahead, how could voice AI actually transform the future of telecommunications?

Nilesh Parikh, Mavenir (03:48):
I think this is a pivotal point where telco is to leverage voice, which has been there, and use it for providing advanced solutions. So we are doing language translations, we're doing accent translations, we're now using speech solutions for defect detection, for providing fraud detection. But in future, I think we are trying to automate that. Automation means during a call — let's say you and I are talking, Tony, for an hour and we need to order an Uber and I need to catch a flight — I can do that during a call and I don't have to go outside of it. And I can natively now use my voice through my operator's network and deliver services. So by automation I'm talking about basically now you don't need Siri or Alexa for some of the commands that you can do through networks natively using your native dialler.

Nilesh Parikh, Mavenir (04:50):
And that's the experience that we want to bring to the market.

Tony Poulos, TelecomTV (04:53):
Well, if I could have access to what I just saw demonstrated, where I can talk in any language and hear the responses in the language I'm speaking, and talking to anybody in the world — that is a massive move forward. And I would think telcos would be jumping at the bit to get this technology and keep their customers happy. I think it's a great start. Both of you, Shawn, Nilesh, thank you so much for being with me today and great luck with the new partnership.