Tom Arbuthnot: Hey, everybody. What has become a bit of a tradition now, we're doing a live on-site podcast from Commsverse. Rishi, who many of you will probably know, we've been talking about doing this for a while, and it's taken getting in person to do it, so appreciate you taking the time. For those that don't know you in the community, Rishi, just introduce yourself and your role at Logi.

Rishi Kumar: Absolutely. So I am the Director of Alliances and Go-To-Market at Logitech, and one of my primary responsibilities is to oversee our partnership with Microsoft across co-innovation, co-marketing, co-sales, essentially every aspect of go-to-market.

Tom Arbuthnot: Awesome. And, like, we're definitely gonna talk about some of the stuff that Logitech are doing, but you have a really interesting perspective on the industry. You're pretty plugged in. Every time we have an AI conversation, you bring something up for around that. But also Logi have lots of different customers on lots of different platforms, so you've got a perspective of what everybody's doing, not just Microsoft. And with that in mind, what are you seeing in the kind of AI landscape?

Rishi Kumar: Yeah, so, you know, firstly, it's an exciting time to be alive. You know, we are in, what? The fourth industrial revolution, and we have Frontier Labs IPO-ing, and there's obviously some hype there. You know, on balance, it's also graduation season, and we're seeing a little backlash when it comes to the commencement speeches. So, you know, there's a lot to look forward to. There is, I would say, also justified anxiety, but the truth is always somewhere in between. I mean, we all believe that AI is going to make our lives better. For those who have 9:00 to 5:00 jobs, you know, clearly there is a lot to gain from the productivity aspects of AI. And, you know, there've been some exciting developments. You know, we've had upstarts like OpenClaw that have really opened everyone's eyes to what is possible.

Tom Arbuthnot: Yeah, in like six months as well. Like, it feels like every... less than a year there's something happens in AI. We went from kind of basic models to reasoning, to the claw models. Like, there's still a lot of innovation happening.

Rishi Kumar: Exactly. And if you're everyone, and especially if you're in IT, for you it's about how do I bring this at scale to my organization? And so we at Logitech, of course, work with all the leading video conferencing platforms, and Microsoft is one of those leading platforms, so they have Teams and Teams Rooms, and of course we'll get into their AI offerings as well. But what we've seen is that clearly meetings and the recap of meetings has emerged as one of the top use cases for AI.

Tom Arbuthnot: Yeah, I think Microsoft have said multiple times, like, that is the biggest use case in terms of compute and usage and adoption because everybody has that problem with lots of meetings, and everybody wants the summaries and the insights.

Rishi Kumar: Exactly. And it's one of those things that doesn't require, as an end user, there isn't a steep learning curve. You just hit record or you go back to something like Copilot, or maybe you have a Teams Premium license, and you just get your beautiful recap. In fact now they've even rolled out video recaps, which I personally love because in two minutes you can get a recap of an hour long meeting, right?

Tom Arbuthnot: And custom recaps as well, so now you can set your own prompts 'cause I actually like, I don't deal well with walls of text, so I want short bullet format. That's the best thing I get. So customizing it per user is really powerful.

Rishi Kumar: There you go, right? So it doesn't require a steep learning curve. You start seeing the value of AI immediately. But again, coming back to the IT owner who has to deploy this at scale, I think one thing that has been a realization for me is that it's a daunting task, right? So, you know, there have been other... You know, Anthropic came up with Claude Cowork, which is also super disruptive, but a little more safer, I would say, for an organization that has been exposed to OpenClaw, right? Like OpenClaw might be great for an entrepreneur, no stakes. But for an organization, it might be Claude Cowork. But then you take that one step further. With Microsoft, the approach, you know, they've got what they call Work IQ, which is, you know, they have established a data boundary with you. It's your emails, it's your documents, it's your PowerPoint presentations and spreadsheets, and there's some institutional knowledge there, and they build that trust with you. They're working with you on compliance and all those things. Whereas if you go with an approach such as, say, Claude Cowork, you're going to have to rely on connectors. So as an IT individual, you have to, you know, think about the trade-offs. And I think that's why what Microsoft's doing is very, very interesting because they're trying to bring the best of these emerging AI technologies into a paradigm that's readily scalable for you as an IT person.

Tom Arbuthnot: Yeah, I completely agree. It's really interesting. I've never seen Microsoft move so fast. So, like, Cowork is a great example. Like, Anthropic did it. Microsoft partnered with Anthropic. Shortly later, we had Cowork in the M365 suite, and now it's GA in months. And OpenClaw, they've got their version of Scout in preview. They are moving at an incredible pace. But to your point, they're trying to do the enterprise safe, sanctioned, secure version of that innovation.

Rishi Kumar: Exactly. And you know, when it comes to deploying at scale, I think IT folks will realize that they're best served through especially if they're in an M365 environment by adopting what Microsoft is bringing. And I believe they've now packaged it in what they're calling the E7, the Frontier Suite.

Tom Arbuthnot: Yeah, that's their suite with the Agent 365, with the extra range of tools. But it's interesting how, again, how fast it's all moving. And I hear the same thing in enterprise. Like, you'll hear individuals get excited about ChatGPT or Anthropic. Microsoft's job is to kind of keep up with that level of excitement, but have the IT pro enterprise safe compliance version of that technology.

Rishi Kumar: And then there is also the spectrum of use cases. So clearly, you know, on the productivity front, we talked about meeting recaps. Hey, very limited learning curve there. But then when you come to agents, right, there is some work to be done before individuals start getting comfortable with the idea of creating their own agents and deploying their agents and learning to trust agents at times working autonomously. So again, I think that's where this boundary of trust that is created in the Microsoft 365 environment with their tools is very, very beneficial. You know, there is this now the new role of the forward deployment engineer, right? Which is needed because it's not so easy to go create the infrastructure, the right connectors, the right use cases, and there are some risks, which is why you do need a role like this to be deployed with the company. But I think it becomes all that much easier if you're doing it within a trusted data boundary. Right? So I think that's, again, something that IT folks, they can take solace in the fact that, okay, I'm creating some autonomy with agents, but at least I'm doing it in the secure boundary that I'm comfortable with.

Tom Arbuthnot: Yeah, yeah. With the vendor I've already said I trust my data with.

Rishi Kumar: Absolutely. Yeah. And then, you know, it's an exciting time for Logitech because from our perspective as the leader in both video conferencing hardware and personal workspace solutions, we... you know, we've had a long and fruitful partnership with Microsoft. It's been over 40 years since the days of, you know, creating peripherals for Windows-based PCs, and now in the recent 10 years with video conferencing.

Tom Arbuthnot: Yeah, you were right at the very start of the Skype Rooms, Teams Rooms journey, and, yeah, as you say, leading player in that space today.

Rishi Kumar: Yeah. And what's exciting is that we get to be the eyes, the ears, and at times, you know, the brain, which is, you know, with all the edge compute. And there are so many use cases still to be solved for. So Ilya had his keynote at Infocomm last week where he talked about Facilitator room skills, right? So this is kind of a call back to a time when you had a host in video conferencing spaces to make sure it was tidy, all the equipment was working. And now you can start bringing some of those skills to the meeting room by virtue of the fact that, you know, you can use the camera sensor, you can use the audio. Logitech creates a sensing device called Logitech Spot, which can sense, you know, the temperature, the CO2 levels, and a few other environmental characteristics. So that also allows things like understanding, "Hey, what is the state of the room? Is there a better meeting spot? Is this meeting space being utilized as much as it could?" because it also has presence detection. So I would say it's just an exciting time for us because there's so many use cases to be solved for. And, yeah, just excited to be partnering with Microsoft. And the last thing I would say is that, you know, Microsoft has been quietly winning at AI. There's a lot of hype around obviously some of the newer companies and the frontier labs. And I would encourage everyone to look into a Wall Street Journal interview with Satya Nadella, I think it just came out a day or two, where he talks about how, you know, value is gonna be created in various areas. And we obviously can't allow some of this hype to take over. And then Brad Smith, their president, also published a recent blog that was very interesting. It takes kind of the more humanitarian side of like, "Hey, how should we be reacting to AI, and how is it gonna enrich our lives?"

Tom Arbuthnot: This is what I mean. You're really on the pulse of what's actually going on and have a hand in it. And you mentioned Work IQ. I wanna come back to that. 'Cause that feels like Microsoft's secret sauce. So when you say Microsoft are kinda quietly winning, like that data layer, and because they have Teams, and they have Teams Rooms and they have Places, it allows Logitech as a partner to, like you say, be the eyes and ears. I don't think many people will have seen the Ilya keynote yet. We did a speed edit of it. We'll link to it. But, like, the idea that the room, the camera room, and the speakers can be like, "Is the room tidy? Is the room prepped? Is the room working?" But even the AI starting to be able to contribute. So, you know, how do I join a Webex meeting from the Teams Room? He showed the idea of holding the phone up with the meeting coordinates. The camera just being able to be like, "I've got the meeting coordinates. I'll join the meeting." And it's that partnership of the Microsoft and the partner ecosystem that brings that to life.

Rishi Kumar: And it's exactly use cases like this. So first of all, Microsoft is solving for these enterprise use cases. Also their approach is very interesting, right? Because clearly, you know, they have partnerships with OpenAI, they have partnerships with Anthropic, so they're taking a multi-model approach, right? So they've hedged their bets. They're obviously developing their own models and then some small language models, et cetera. And again, you know, what is interesting is that there's been a revelation in the industry, especially with Anthropic doing better these days, that, oh, wow, the money to be made is on the enterprise side. Well, Microsoft knew that all along. And that's why I say they've been quietly winning. You know, I don't remember the exact numbers, but if you look at their latest earnings report, it is quite astounding how quickly their AI business has grown, which about two years ago was close to zero, and I think it's maybe north of 50 billion. Don't quote me on that, but check it out for yourself. So what seems like a revelation for OpenAI and Anthropic has been what Microsoft has been very quietly doing really well, because they've always been for enterprise. And that's where now Anthropic and even OpenAI with their, I believe it's called deployment company, are learning those hard lessons.

Tom Arbuthnot: Yes. I feel like it's a lesson that companies learn time and time again. They're like, "We'll build this amazing product, and obviously the enterprise will understand how to use it." And then you see this three years later, they're like, "Oh, turns out we need a load of consultants and business change to get people to understand."

Rishi Kumar: Exactly. 'Cause the value is clearly there. And, you know, unlike software where there's a huge demand side economics, you create once, and then every time you sell software there is no incremental cost. With AI, there is a very significant incremental cost, right? And of course all of these companies are infrastructure companies now, because they have to stand up their own infrastructure. So essentially AI generated memes and images is just not gonna do it. You cannot be a profitable, successful entity doing that. So that's why I keep going back to Microsoft has been quietly winning. They've had the right strategy since the start. Yes, there's hype out there, but if you're essentially an IT and an M365 shop, you're very well-served by exploring the solutions that Microsoft's bringing to market.

Tom Arbuthnot: And what's your thoughts, Rishi, about AI on the edge or in the room? 'Cause it feels like that, certainly look at what you're doing with multi-camera, we're gonna have more AI doing clever things on the edge, be it direction, be it camera control.

Rishi Kumar: Yeah, so sticking to the theme of quietly winning, Logitech has also quietly been winning with AI on the edge because we've been working with AI back when AI was sexy or mainstream, right? So, you know, simple things like we launched a product called the Rally System, I wanna say close to nine years ago now. And you could do up to seven mic pods in the room, and each of them had eight beams forming, and sub-millisecond accuracy for noise cancellation. All of that is happening because we put AI on the edge. So we have a lot of these models that are running on the edge. We're doing the same with computer vision. So of course, you know, we understand that the cloud is expensive. If you're following the news on all these infrastructure build-outs in the US and other places, it's really expensive. There's a lot of energy requirements. So the more you can put on the edge, the better. And some use cases just make sense. You don't wanna do a round trip when you can do certain things in sub-millisecond accuracy.

Tom Arbuthnot: Yeah. I did a, at the bootcamp, I had a session with Irina, which we put on the site, and we were talking about that. She owns the rooms and the room management, and saying how much value there is of having AI on the edge. Certain things that are just better to do on the edge and certain things are great in the cloud, and it's gonna be a partnership of picking the right things to do and where.

Rishi Kumar: Absolutely. And then that's why we are so excited about working with Microsoft because there is a plethora of things that still needs to be done, and then some of it is gonna be done on the edge, and clearly some of it, you know, there's value doing it on the cloud.

Tom Arbuthnot: Awesome. Rishi, thanks for catching us up. I'm glad we finally got this on the record. And, yeah, appreciate all the support of the community and the partnership as well.

Rishi Kumar: Likewise. And appreciate what you do because everything's moving so fast, so you know our audiences need to be kept up to date on what's new and what's worth looking at.

Tom Arbuthnot: Awesome. Thanks so much.

Rishi Kumar: Thank you.