Ray Le Maistre, TelecomTV (00:05):
Artificial intelligence is having a dramatic impact on the telecom sector that isn't just an opinion, it's a fact. And to get some insight into that impact, I'm talking today with Danielle Rios, CEO at TelcoDR and acting CEO at Totogi. Great to chat with you as always, DR. And great to have you at TelecomTV's DSP Leader World Forum, which kicks off today. Now, as you well know, the pace of AI innovation is relentless. As someone who has always been on the forefront of public cloud and AI developments, what recent announcements have caught your eye as we kick off our DSP leaders event?

Danielle Rios, Totogi (00:45):
Yeah, you're right Ray. So much is going on with AI and as I was thinking about preparing for the DSP leaders event, I was like, let's look back just one year and see what was going on with AI in 2024. And so back then, which wasn't that long ago, the large models could only sustain short, relatively simple tasks and offered little autonomy. And so now in 2025, the new releases can now work for hours on complex projects, handle much larger data sets and deliver best in class coding, which is really great if you're a software provider like me. So for example, last month, philanthropic dropped Claude Opus four and Sonnet four. Opus four can independently tackle complex multi-hour tasks like coding projects or in-depth research. It outperforms previous models in both reasoning and programming. They now claim world's best coding model title with benchmarks showing it beats the other models in coding and tool use.

(01:46):
OpenAI released GPT-4 five and New agentic models 03 and Operator. It has improved emotional intelligence, longer context handling and the ability to use tools and can execute multi-step workflows autonomously. Not to be outdone, Google is still in the race with Gemini 2.5 Pro a frontier model with a 1 million token context window advanced, multimodal reasoning and agentic capabilities. And so it enables AI processes on vast data sets and complex tasks. And this is building on the other really big announcements that happen at the end of 2024, which is model context protocol from ANTHROPIC or MCP, which gives you seamless, secure and universal integration between AI models and external systems like say, enterprise systems like BSS systems, which can dramatically simplify how AI applications access and interact with real world data. And so it is rapidly advancing. Every time I feel I can log onto x.com, something new has been dropped and you got to go check it out. And so to Totogi, we're constantly looking for ways to use all these new features in our products

Ray Le Maistre, TelecomTV (02:58):
Product. Yeah, it is just incredible how much change there is every week, every day almost. And of course you are right at the heart of this and many will already know that to Totogi launched BSS Magic, which as I understand it isn't a BSS product as much as something that adds an AI layer on top of a telco's BSS estate. How are you using the latest large language model capabilities in BSS magic?

Danielle Rios, Totogi (03:28):
Yeah, every time a new feature comes out the Totogi team, they're like kids in a candy store. We just get on it, test it out, are the claims real? Can we really use it? Is it as great as everyone's talking about? And then if it's great, we totally use it. It's awesome. And so I just mentioned model context protocol or MCP. And so when you combine this with our telco specific ontology that we built, it really accelerates how quickly we connect to these external data sources, right? Like say a BSS product made by another vendor. And we can start changing how your business processes work. So we love MCP, it saved us a bunch of time encoding in our own product and we absolutely love it. Next, all the agentic capabilities being a released are also helping us to accelerate what we can do for telcos in our own product.

(04:18):
We demoed this at MWC and we'll have it at TM Forum's, DTW Ignite event, which is in Copenhagen in a few weeks. And so really excited to show people how this product is advancing. We really do need to keep up with it on at least a monthly basis because it is changing that fast. Every time something big comes out, we test it, we incorporate it, and we move on. And so this is a very dynamic product. I think it's one of the leading AI products in Telco and it actually does work and we're really excited about it. And so we're piggybacking like crazy on all these model enhancements and rapidly changing our product to be at the forefront of ai.

Ray Le Maistre, TelecomTV (04:59):
And as you mentioned there, this isn't a plan or an idea. I mean, this is something you've already got out there in the market. And to Totogi now has several tier one customers using BSS magic to solve specific operational challenges. Can you tell us about the Asian tier one telco that's using BSS magic with cloud sense to make their sales representatives more efficient?

Danielle Rios, Totogi (05:25):
Of course, this is a really fun idea because it came directly from the telco. This wasn't a sales cycle where we were pushing different ideas, but this was direct from the telcos itself. And so it was a cloud sense customer that we all know, dealing with sales reps is always a difficult thing. Train them on new products, sometimes they don't listen. And so they really wanted an AI agent that took over portions of their sales cycle and do it the way the company wants things to get done. And so we built an end-to-end agent for them and the results were amazing. They challenged us to use BSS magic to cut the time for their sales team to quote an enterprise sales order by 15% so the sales reps could spend more time with customers. And so that's what you want your people to do, spending time selling and not really fighting with the software.

(06:15):
And so in a few weeks, the agentic flow was up and running a production and their minds were blown. I think the direct quote was like, wow, this really does work. And so we demonstrated high business impact in a very short period of time that's resulting in faster quotes, higher quality results, and were creating actual real business value for the telco. And so I think one of the problems that we see is people are overdesigning their ai, they're kind of taking a big bang approach and trying to do too much all at once and starting with a project that's a little bit too ambitious. And so what we're seeing as being very successful, starting with something small, getting into production and expanding from there. And there's so much to learn with ai, not just the own IT teams around these projects, but the users themselves. And so we think this is the best approach to get real results.

Ray Le Maistre, TelecomTV (07:09):
I mean, that makes a lot of sense. Now, another tier one operators using BSS Magic to solve technical support issues. What can you tell us about that use case?

Danielle Rios, Totogi (07:22):
Yeah, so this is a different example. Everyone knows IT systems have bugs, and a lot of times that burden goes to the technical support team within the IT group to solve the problem. And so this is a Vera customer that was seeing an unusual uptick from a different system. They were seeing four times the normal support tickets. And so they came to us and said, can you help us and provide more people? And I actually said, Hey, can we solve it with AI maybe? And they were open to that idea. I'm like, instead of throwing bodies at the problem, which is usually what we do, that's kind of the old approach, pre 2022 right now, I'm like, Hey, maybe we can use AI to do this. And so it had a lot of eyes on the problem. It was blocking revenue. Stakes are high, right?

(08:07):
Temperatures are really high internal. And each one of these tickets to solve manual intervention and on average about five hours to resolve. And so we applied BSS Magic and AI again, we created an Agentic Flow to review the tickets and resolve them. And we brought the time to resolve each ticket from five hours down to about 15 minutes. And so we didn't need more people, we needed fewer. So we still had humans in the loop to make sure that AI wasn't crazy and was doing the right things. But AI would do the work. The human would review it and say approved and incentives on its way. And so again, we could build something very quickly, solve the issue and unblock revenue. And so it has high business value. And so now this telco is thinking about, Hey, what else can you do? Can you add it for these other systems and help us really squash our tickets and support really speed up that resolution time.

(09:03):
And so what's so awesome is that as the ticket volume goes up, it's really easy to clone in AI agents, literally a copy paste. And so instead of adding more bodies to solve the problem, just add more AI agents. And so we could have 10 AI agents and then take it down to four one whatever you need. And so it's kind of this completely different approach to solving problems that we're seeing. And I think the telcos are really surprised. It's so easy to add it and get it up and working in such a short period of time and have impact. So it's awesome.

Ray Le Maistre, TelecomTV (09:36):
Okay. I think this is what everybody's looking for. It's real world use cases to build on. Yeah, for sure. Now also, you recently published a blog post about how to Totogi is doing something that vendors don't normally do, which is helping a customer leave their platform. Now many vendors might let Telco fend for themselves in the hope of extending the engagement and associated revenues, but why are you taking a different approach?

Danielle Rios, Totogi (10:07):
So as you know, I acquired CloudSense a little over six months ago and we encountered a customer, a tier one customer based in Western Europe that had been trying to leave CloudsSense for three years. And the trickiest part had been the data migration really mapping the old data structures from the cloud SIM system into their new vendor. And you are never happy to see a customer leave, but I had this idea, which is this is a large tier one with a significant BSS estate. If we were able to show them how well BSS magic could work in this scenario, they might become a Totogi customer with BSS magic and actually spend more than they were spending with CloudSense directly. So I saw this as a really great opportunity. And so I think this is another example where BSS magic, again because it works against any vendor system, it's not just the ones that we're building and providing, but really this AI layer that you're adding and tackling one of the trickiest parts about swapping systems or even upgrading systems, that tricky migration when there's a data schema change, AI and BSS magic are really, really good at this problem.

(11:26):
And so we're applying it in this way. And so I think there's tons of opportunity across Telco for people who feel trapped by their previous vendor or it's going to take years and millions of dollars to move off. And again, the incumbent vendor is very reluctant to assist and the new vendor is just itching to get that data into their new system and start getting them live to Totogi is super happy to help with this. And so it's another situation where I think again, demonstrated three different examples of accelerating revenue in a live system, helping with support tickets on another system, and then helping move data different ways that we're applying AI and BSS magic in the real world to help telcos advance their business.

Ray Le Maistre, TelecomTV (12:13):
And we're going to be talking more about this kind of thing today, in fact, because you are participating in a panel here at DSP Leaders World Forum on leveraging AI throughout the network that we're going to be streaming live and then making available on demand after the event. Now, if you had just 30 seconds to give advice to telco executives about using AI in their network, what would you tell them?

Danielle Rios, Totogi (12:41):
Yeah, I think again, this is not a big bang approach where you need to put it everywhere in your telco immediately. And so I think what is really working and we're seeing really great success are the telcos who are willing to start with something small but impactful. Start with a real problem. Your sales reps are not compliant, right? Too many support tickets or a migration that's not going well, right? Get it, get something small, get it deployed, and then start expanding and take on new use cases. We've found that this works. You will too. And I can't wait to have a super lively discussion about this. DSP is always amazing, and I'm sure this year we'll be no different. So really looking forward to it.

Ray Le Maistre, TelecomTV (13:26):
Okay. Excellent. Well, I'm looking forward to it as well. Danielle, thanks very much for talking with us today and enjoy DSP Leaders World Forum.

Danielle Rios, Totogi (13:34):
I will see you later.