AudioCodes CEO Shabtai Adlersberg - The AudioCodes Journey, Hardware to Saas to AI

I was always dreaming of building a voice moat, right? You know, a moat you want to build for yourself, a fortress where nobody can get near you. Just like big companies, great companies like Microsoft, Cisco, Check Point, they were first, they excelled in building a very strong business and then it's very difficult to come after them. So we're trying to do the same on the voice side.

Thank you, Tom. Glad to be here with you this morning. Well, as many of you know, I'm a co-founder of the company. Go back 31 years ago, co-founded it. Been an engineer, a development engineer in the voice space, and I'm CEO and president since then. I also served as the chairman for like 20 years, but then they pushed me a bit. Help, never mind. So I'm an electrical engineer by education, and I've been dealing a lot with digital signal processing work back like 30 years ago, and then dealt with system architectures. So I really love that, that's why I'm still in that business. I love the work, I love to see great new design and new solutions coming out. What I love most about the industry and AudioCodes is really the strategy play, because as a CEO of a company, you really need to chart the company's way through changing markets, trends, new competition, new technologies, new applications. And as we all know, take any successful technology company, if you don't reinvent yourself every five, seven years, you're out of the game. I see my key role as CEO really is to navigate through those times and always try to, while we work on developing, marketing, and selling the current offering, we always need to look forward, be as they say in French, clairvoyant, right? See for the future. So that's why I love this game. I love the game because of all these strategy games. Been around in the industry for like 40 years now. Prior to that, I've spent like nine years in the Israeli Army as an Intel development engineer. Prior to co-founding AudioCodes, I've co-founded another public company which was acquired like a year or two ago called DSP Group. So DSP Group was developing chipsets based on DSP technology, and it was fun and wonderful. So long career in what I would say the voice industry. And then personally, I'm a happy father to four great children. Two elder sons, one which works at AudioCodes and helps me on navigating the new business, and one that is a psychologist. And then I have two little daughters, and they have been wonderful in bringing joy and light into my life. So pretty happy family around me, and I should also mention that I just became a grandfather like four weeks ago.

That's a lot, and it's fun. You want to be in the action with the family around you, the business, that makes your life interesting. That's who I am.

Right, it's all about passion. It's all about passion in dealing with what you love to do, and sitting on top of it always, as I described before, looking around me to identify what's going on and trying to see ahead. So yeah, we are a company that, as you have mentioned, we went up the food chain. We started as two algorithm guys. My co-founder Leon was exceptionally involved in voice compression technology. I was a good engineer, but then we decided, okay, Leon takes R&D and technology, the rest of the company to take care of. So we went up the food chain from algorithms to chips. Algorithms was '92, '94. Chips was '95, 2000. Then we tried to sell chips to the leading DSP board companies at that time, names like, or older people here on the call, names like Dialogic, Natural MicroSystems, Brooktrout. They didn't want to buy chips, all they wanted to do is to license the technology. And we said, well, we won't make much business that way, so we want to build our own cards and boards. And we were lucky to become in two years a leader in the industry. So that was '97, 2000. Then that was the era where gateways were built out of cards that were plugged into a PC. So names like VocalTec, VocalTec is the originator. Then it was obvious that in order to optimize the product, you need to go to a carefully designed and built company gateway. So well, lucky that happened, side by side with the internet bubble crisis, so we had enough time to build our own gateways. And soon we became a leader in the gateway market. Put aside Cisco which has a huge captive market itself, we became number one in the gateways. So that's between 2002 and until today actually. And then we did some acquisitions, some that worked and didn't. We were acquainted to the SBC world in 2006. There was a little startup in Dallas, Texas. So we thought we would ride with it. We were wrong, because I'll tell you what's the reason. It's a good lesson for anybody that listens: when there's a new market developing, always go with the least complex product. Always go from the low level and work your way upward. And the startup, and this was a pricey lesson, the startup we bought was building a big, huge SBC monster which obviously nobody had any use for that when the market happens. So that was not successful. But then we decided internally that we would go and build our own technology, which started back in 2008 or '09. And since then we became a leader. If you look to the market studies, you'll see that in the Enterprise world, because we do not target so much the service providers, we became number one. We're holding a market share of like 23% in Enterprise SBC. Oracle is like 20%, and all the rest are below us. So those were the SBC. But then we came to the conclusion that we really, and it's actually tied up to Microsoft, we made a strategic decision back in 2000. We didn't know it was that strategic, right? But I was going to ask this, how much vision you had as to Microsoft? They always come to disrupt the market and they always say they're going to disrupt the market. You know, I have a funny story for you. So I'll take you backwards. In my previous company in '92, so we were voice guys, we were listening to ITU and standards. And then leading companies were names like Lucent and Alcatel and Nokia and Ericsson. And I was a fan of the Nokia phone. And then we have been approached by Intel. Intel was then working hand in hand with Microsoft trying to start some work into the computer world with voice. And people came to me and told me, hey, listen, Intel wants to... and I said, come on guys, are you crazy? You've got those huge giant companies like Lucent and Nortel and all this, you think that Intel will be able to... well, I was wrong. And by the way, Microsoft was also part of that wrong. So I'm not the greatest strategist if you will. So 15 years later, I knew better than to not pay attention to Microsoft. So when Microsoft came out with OCS, Open Communication System, and at that time Cisco and Avaya were the leaders in that world, I said, well, you know what, they are coming with an application, they'll be fed, and they have the pockets. Avaya already then was not at its best times. So we took a decision to stick with Microsoft back in 2010, and that is making our story until today and I guess for the foreseeable future.

Yeah, and quite frankly, we've been putting most of our weight behind that effort. And we were like, well, I need to give huge, huge thanks to our people because I can be the greatest dreamer in the world, but if I don't have a good, strong R&D base that helps me deliver that, and they came out with wonderful solutions up until today. So we became fairly fast. We had a tough competition at that time, there was a company called SS8 Networks which afterwards became Ribbon Communications. And they gave us a fight, they gave us a fight. But back in 2015, when they had a little crisis, they decided to back off Enterprise and go to service provider, and we took advantage of that situation and moved forward. But the more important step was, as I've mentioned before, going up the food chain, was that we said, well, enough is enough with selling best-of-breed products. Okay, we could be the best gateway and the best SBC, but we do not control the solution. So we then decided to come up with the concept of One Voice. And One Voice said to Microsoft at that time, Skype for Business customers, it told them, hey guys, we bring you the majority of the components so you don't need to run between different suppliers. Take a gateway from that one, SBC from that one, a phone from that one, management from another company, recording... so we said we'll do it all. And apparently that was the beginning of the most important step we took. So we started to win more. And by the way, we just heard about two years after we made that move, throughout the grapevine, we heard that the VP of marketing of Sonus said, well, that was the move that killed us, and we were not paying attention to it. So we came up with a great solution. And through Microsoft evolution, so it was Skype for Business, then together with Microsoft we invested in Skype for Business Online, which was not the greatest product on Earth. They backed off. And then the competition, there were steps in the industry, Slack gave Microsoft a run, so Microsoft had to come out with Teams, which initially was a disaster to us. I said, I go for a new solution, we have Skype for Business, it's a great solution. Apparently, yeah, we were wrong. And thanks, we then, so we joined, if you can't beat them, join them. So we joined the Teams camp.

Yeah, and the product was not mature. I think we got on the bandwagon somewhere in '18, '19. The product was not cooked yet. I think we all enjoyed, well, enjoyed is a strong word, we did not enjoy the pandemic, but the pandemic really gave a huge boost to all that industry, right? Working remotely and working from afar and being able to maintain meetings and discussions and contact center agents sitting remotely. So I remember the numbers, right? December '19, Zoom had like 10 million customers and Teams was like 17 or 18 million clients. Go forward six months, Zoom boasted 200 or 300 million users and Teams was always ramping up fast. So that was the evolution. And again, so if you ask me about AudioCodes and what's the main business we have, well, it's all centered around UC and some CC. I'll talk about the future in a minute, but it's mostly UC. And in UC, we attach ourselves to Microsoft, which these days holds like 40% market share, etc. And we say, so that's the strategy. Strategy is stay strong on Microsoft Teams arena and build yourself always upward, just not to leave any chance to your competition because they will have to build all those layers. So it was then that we started to build the Live business. Live means our managed services, because at the end of the day, Microsoft Teams is a pretty complex solution and for large multinational companies, it's really tough to get into it. You need to hire professionals all around the world and it's costly, etc. So we are successful with that offering. We're glad that we started it back in 2020 and now this year plan to end up with a business that's like 65 million annual recurring revenue.

Okay, well, I can tell you that sitting in management meetings back in let's say 2012, when you were a great voice networking company, however we scratched our head and it was really tough for us to understand how we're making that transition to the cloud business. And it took us several years. But I can tell you that I'm amazed. Take the Live platform, and the center of our success with Microsoft is a new platform that we introduced about a year ago called the Live platform. So Live platform is really a combination that starts with connectivity at the bottom with gateways and SBCs. You add on top that management of devices and users and sites, and then you do some integration. So that is the basis and it makes our deployments much faster because otherwise it's a long, tedious project. So Live platform is that. Now the key, so we call that strategy land and expand. So we are capable of landing in various end-user accounts through our superiority in the connectivity business and management of users, but then we're coming up with a set of value-add applications. Right now think about GenAI coming to service with conversational AI. So we're gradually adding more and more services to that Live platform which again will fortify and solidify our... I was always dreaming of building a voice moat, right? You know, a moat you want to build for yourself, a fortress where nobody can get near you. Just like big companies, great companies like Microsoft, Cisco, Check Point, they were first, they excelled in building a very strong business and then it's very difficult to come after them. So we're trying to do the same on the voice side. And so those are the value-add applications.

For us, it's a blessing. It's a blessing simply because we see it, it's our glasses, probably other people would see it differently, but we see voice at the center of both. Because your communication collaboration solution that serves your employees may then help you, you want to converge the contact center. Until today, I think voice is like 50% of the traffic in contact center area. So you better have a cohesive solution that takes care of the same technology components and serves the two. So we're very, actually we are among the first. I think we had a great Enterprise Connect show earlier in March this year. So yeah, we think that that convergence is happening. And obviously companies would like to have one combined solution. And that's all again playing into this Live platform play. So you can offer voice, you can offer recording, you can offer contact center, all on the same kind of platform to the customer. And now we're adding conversational AI to the mix which further elevates. Well, throughout the years, if you want to be competitive, the only way to be competitive these days because there are many other companies is to be able to penetrate and I would say conquer a little hill, but then add to it, integrate it, make it bigger and bigger and bigger. As you go bigger, those who come with some very brilliant solution next to you still are not capable of bringing a complete solution because it takes years and hundreds of people. So that is a lesson that we always, once we identify a threat or a product, we'll think about, okay, how can you, I hate the word imperialism, but then how can you build an empire around that? And that's the approach.

Okay, so it all started with me sitting daily in six, seven, eight meetings a day, 30 minutes long, and getting out of those meetings you don't remember what was decided, what were the action items. So we started to develop it prior to the pandemic, and it was hopeful because the voice systems, the recording quality, the meeting rooms were awful. So it was really a slow start. We really struggled with all kinds of issues. Then came the pandemic which drove a lot of the collaboration solutions like Teams and Zoom, so that made the system essentially better in quality. It was spread. And now Microsoft with their meeting room initiative and Zoom, they all want to make your meeting room the center of the business and decisions. So we were able to evolve from that initial product. The idea, listen, there are two levels, and basically here is something that I don't think many people paid attention to, but we are not the first note-taker in the industry. You have a series of startups, names like Otter and Fireflies and others, which are doing it mainly more in the Zoom space. And then there are the big guys, you have Zoom and Microsoft who are also developing their own first-party. Now, so I'm building the product as follows. We try to excel on the meeting level. So we take care that recording and transcription, by the way, transcription is not an easy thing simply because yes, you can get 80, 90% accuracy, but once you get to different organizations, one from the sports arena or from government or politics or healthcare, the huge amount of concepts and terms that they typically, speech-to-text will not recognize. And arguably the most important terms for them and their company, right? They are the terms that decide whether we, they, or what they meant exactly. Now it's a chain, and the strength of the chain is, as you know, related to the strength of each of the little pieces in it. So if you don't have a perfect start, your inference, your LLM solution and prompts and everything comes afterwards will suffer because it will take different terms. So we pay a lot of attention of course to come up, and we just launched a SaaS solution back in March. We're happy these days, we have already tens of customers, about 50 customers, growing fast, more than 10, 15 a month. But that's on the meeting level. But then, myself, I consider myself also not only an engineer but a manager. And to me, the fact that there's an important meeting taking place today in San Francisco with an important customer, and I, the decision maker, have no direct access to it because yes, I'm told, yeah, it was a great meeting, yes, they are interested. At the end of the day, like in any other step in the industry, you need to validate. And once you get access to that recording, and of course all those recordings, three years ago nobody in the industry wanted to hear about it. Nowadays they ask you, well, you have a recording of the session? I'd like to get it. So you get the recording, you get to understand who said what, speaker recognition technology, who said what, and then you get the insight for, I'm a project manager, you are a sales manager, we have different perspectives on what's the summary of that meeting. So we're getting all those recaps. And now we're putting them all at AudioCodes because these days we have about a thousand employees and about 500 are US-based. So in that environment we have today close to 200 meetings a day. All those meetings, recording, transcription, and insights, they all go into one repository. And now management can assign small teams that focus, let's say, on SBC or IP phone or recording. We can identify, we attach tags to every meeting so we can pull out a bunch of meetings relating to a certain topic. And then I can work with, guess what, the best chatbots in the world, could be Copilot, could be Q, could be Gemini, could be Claude. And now because these guys doing nice work but not great work, right? I think everybody that's been using those chats know that it's accurate up to a certain limit. Why? Because they do the search on a huge, humongous amount of data. If you narrow it down and take now Meeting Insights as a preprocessor who identifies those X meetings that relate to it, now the chatbot will do substantially better work. So that's how we're trying to position Meeting Insights.

Correct, correct. Yes. So to me, Meeting Insights now is not only about meeting insights, but it's also about business insights and it drives better decision making. Also drives knowledge management. Think about the day that one of your best employees comes and says, hey Tom, I'm leaving, I got a better offer, going overseas, and you're losing a lot of knowledge. However, think about having the company records of this guy and everybody's meetings for the past so many years. Now you bring a new guy, you know.

So Meeting Insights just started a huge, huge journey into the future. And we're adding, we're talking to customers, every customer brings up a new angle of sight. So it's really exciting. It's really exciting, and I think it's got a great future.

Absolutely. It's like they say, nine ways to skin a cat, so a thousand ways to skin a meeting. And yeah, it's great. And then by now, I'm stepping one step further. Okay, because if you ask me, okay, where is the future of the company? So on one end, we put our bets on the Live business and the Live platform for the UC and Microsoft Teams business on one end. But then there's a huge opportunity that's starting up right now in the conversational AI business. Now here, by the way, there are two ways in which we go to market. One which is very simple, the traditional. Gadi is great and his partner are great in coming up with VocaC, which is a great first solution for Microsoft Teams. So that's a first-leading SaaS application. We have SmartApp, which is a very nice either compliance or interaction recording solution that also is a great SaaS application. And then Meeting Insights. So on one end we will try to come up with applications that each target a different need. But then there are fully, I would say, closed or would evolve, but still it's all surrounding. But what happens now if you go, and we worked here with the Israeli Red Cross organization. So their biggest issue is getting millions of calls every year, hundreds of calls an hour, and they need to resolve what's happening and they need to resolve the most important thing which is where do I send the ambulance? Because if you have two names of cities which are spread 100 kilometers far from the other one, and you know names in the UK doesn't come, are very close, you can think about yourself, Yorkshire and Berkshire, and just think about that, it would be a disaster if you just send the ambulance to the wrong address, right? So to make that solution work, we started with some real bad noisy cell phones. So the accuracy rate of the speech-to-text was like 40%. So we had to connect to telephony, we had to build a good networking solution because it went into a contact center, we had to attach recording to it, we had to put some management solution to control the devices, we had to do transcription. Now we build some inference solution on top of that using LLM technology. Think about how many organizations in this world are in possession of all those diverse set of technologies and tools. We're lucky that we have that 30 years behind us that we have developed all that telephony and networking that now we can throw into the game conversational AI. And now you come up with a perfect recipe. So we put huge effort in developing not only SaaS solutions but also what I would call custom SaaS or custom projects for big companies because everybody needs to customize it for their telephone network or CRM or anything like that.

Right, right. So that's why you hear me excited in this call because we did a lot, but that's a great thing to do going forward. So on one end it's Microsoft Teams and Live business, contact business, and then the conversational AI which is completely new for us.

Okay, well, thank you, Tom. Really appreciate the opportunity to be here today. And obviously, let's stay in touch and let's meet every so many months and we can talk about the new things we see around.

Sure, to do that.