Tom Arbuthnot: Hi everybody. Welcome back to this briefing. Really excited to, uh, have this one. We're gonna talk about, uh, rooms and devices and, and AI, and we are just coming into the, the new kind of financial year as we talk, uh, about Microsoft. Uh, lots of planning has been done. There's lots, lots of exciting things coming along.

Uh, I'm really excited to have Irena on the show. Um, I think first time you've been on possibly as well. 

Irena Andonova: Yes. Yes. I think that is the first time in one-on-one with you. 

Tom Arbuthnot: Oh, I dunno how long this has taken so long. So I appreciate you taking the time. Um, maybe we could start with, um, your current role and kind of responsibilities and then I'd love to take it back because I know you've been working on the product for right since the beginning.

So talk through kind of your journey with, with teams and devices. That'd be great. 

Irena Andonova: Yeah, of course. Uh, well, hi everyone. Uh, my name is Irena. Uh, I am the head of product for Teams devices, uh, at Microsoft. And this is different device categories, uh, Teams Rooms, Teams Panels, Teams Phones, uh, that's all of the Teams under the Teams Devices, umbrella.

Um, as kind of to alluded, I've been on the product area for some time, kind of done different things. Um, I started together with, um, I Stein, which I'm sure many of you know. Uh, if you're listening to this show, uh, in, gosh, it's been a while, I think 2016, uh, when we first came up with the concept of what was at the time, Skype Room System V2, uh, that was the predecessor of what now is Microsoft Teams Rooms.

Of course. Uh, many years, uh, later, and many, uh, thousands, actually millions of devices deployed worldwide. Uh, very, very proud of that. I was the first PM on the project and, uh, a little. Anecdote, uh, if you wish. Uh, we announced Project Rigel as it was called, code named at the time, and the little anecdote is that Rigel is a supernova or a blue giant in the Orion Constellation.

It came up from my daughter's astronomy class who was super excited about it and I was thinking, what should I name this project? It's gotta have a name, name so we can rally around it and hence it became Rigel. It is.

Tom Arbuthnot: Oh, that's great. I've, I've, I've never known the, the origination of that name and obviously known the name since day one.

That's really interesting. 

Irena Andonova: Yeah, it's still in some of the code parts, um, on the developer side. So, uh, pretty interesting. So that's, uh, Teams rooms. I've worked also on the voice services, um, and Teams phones. Um, and then I took a little bit of a detour and I worked on what now is pro management, uh, service and portal.

Uh, that was built by my team, uh, from the ground up. Uh, we were part of the management division and, uh, build a management service. Um, which then later on became the pro management portal, focusing entirely on Teams devices. 

Tom Arbuthnot: Awesome. That's amazing. It must, must have been an incredible journey, particularly the, the, the scaling of where we are now with rooms and devices like it's, uh.

Uh, did you kind of have some feel of the, uh, the potential? Is it kind of blown you away? How's it feel? 

Irena Andonova: It, it is, it is entirely amazing. Uh, and I feel very, very proud for the work the team have done, uh, over the years to achieve where we are. I mean, I think we announced, uh, few months ago. We had at the time crossed a million mark for Teams rooms, and of course there is many other millions of the other devices in the other devices categories.

Um, so it is very. Uh, an amazing growth, uh, so to speak, especially after the pandemic. The video adoption has been very interesting, very interesting across the industry. I think during the pandemic we all learned. Um, how to communicate with video and how to adopt to video communications. And now it has become, because of course, and that wasn't necessarily the case when we started, and it's really great to see the adoption.

Uh, we have reached. 

Tom Arbuthnot: That's awesome. And, and part of the premise of, um, what was Skype room systems and our Teams room systems is it's built on a, uh, compute platform that can be added to and developed over time. So, you know, we got incrementally new features over the lifecycle and it feels like we've hit a really interesting time in the industry now where AI's come along and you've got all this capability to start leveraging and adding AI across the.

Rooms and devices and across management and operations as well. 

Irena Andonova: Absolutely. It is. Uh, it's very interesting time that we live in, uh, at this point of time. I think many of us, probably all of us, uh, have really learned how to use AI in some, in some ways and how AI impacts how we work. Uh, uh. Chat, GPT, Copilot and other technologies have become very widely adopted.

What I usually think about, uh, when it comes down to the devices, um, AI starting to impact not just how we work, but also where we work. Mm-hmm. I often joke that our team is perhaps the only team or a, um, a team designated to think about AI in the physical space, meaning what happens and how does AI help the users when they're in the meeting rooms or when they come to the office?

So many customers have a return to office or some kind of hybrid mode of operation. When people think about like, oh, I'm gonna be in the office and I want to go on certain days, I wanna meet with Tom. So maybe Tom and I are both in the building. So how do I know first that we are both in the building and where should we meet?

AI can help with that. That is the kinds of technology that AI can really help us. AI can help us locate our coworkers, uh, proactively find a space to meet, make a reservation for that space, and then help us have a very, very productive meeting. I also often hear from customers. Especially the ones that have a return to office mandates of how we can make the spaces and how we can make, how we can engage the users better.

How can technology help engaging the users? 

Tom Arbuthnot: Yeah, it's interesting that kind of earn, earn the commute or encourage people to come back and it's, it's, it's a real challenge because actually we learned during the pandemic that, you know, working from, from the laptop. Was quite, quite plausible and, and were well, and then we got AI in Teams obviously.

So I've got my AI in Teams. So actually you, I feel like from a a rooms perspective, you have to match or better that experience. 

Irena Andonova: Absolutely. And, uh, that, that is, that is the challenge. And that's, uh, that's what I'm hearing from customers of what can we do in the meeting spaces. So when we're in the office and we meet in the office and meetings, we, we know, and we, we've known that and we have learned meetings happen in the meeting rooms.

So when the people are there, how can we help people have better and more productive experiences? So one example, a, a very simple example, um, summarization or note taking is something that has almost become ubiquitous from AI perspective, but it is really something that I don't know anybody who likes to take notes.

I don't. And I love it when there is something or somebody who can help do that. So if we are in the meeting space, if we're in the meeting room. Very much like if you are just in a hybrid meeting, the meeting room can participate in the note creation. It can participate in summarizing the key points from that meeting even if there was no meeting scheduled.

So Tom and I just walked in a room and we're having a conversation. Who's gonna take notes? 

Tom Arbuthnot: Yeah, this is really interesting. This is fully in person, so not a Teams meeting, but just like a full in person. But we've got the smart compute there with the, the Teams room. How do we leverage that experience?

Irena Andonova: Exactly, exactly. Another example is, again, reflecting on, um, how AI helps with, uh, productivity and how it helps users. I, I'm sure many of us have too many, too many meetings to attend. So what if I don't make it to a meeting or if I'm late for a meeting? I wanna catch up with, with the transcript, with the summary notes.

In the summary notes. If people are joining from their laptops, you would see Irena was speaking about AI summary, um, and so forth. But if there is a room in the meeting then it's also very important to show who said what, to know who was in the room. So that is something that we have already made available on uh, Teams rooms on Windows and on Android.

Being able to reflect who is in the room in order to properly attribute who said what. That is very, very critical. So we can make the transcript reach and we can help Copilot be more helpful to the user. So if I wanna catch up later and I wanna see who said what, I can definitely do that from the summary notes using this, um, technology, uh, or a skill of speaker attribution properly attributing who said what in the meeting room.

Tom Arbuthnot: And that technology wise, that's a, uh, combination of kind of people enrolling their, their face and, and voice so the system can understand who in the room said what. 

Irena Andonova: Yes, absolutely. Uh, super easy to do, uh, from the Teams client. Anyone can enroll their voice profile. It helps, uh, it helps actually, twofold.

You can do a feature called voice isolation, where if you have a barking dog on the back or a lawnmower, uh, you will not hear that, uh, in the meeting. So benefit there as well. But it also, if my voice profile is enrolled. Then my um, name can be properly attributed in the transcript as well. 

Tom Arbuthnot: Awesome. And that for those kind of, uh, that have been, you know, loosely following this space, like that used to be exclusive to certain devices or MTROW, then it became all MTRWs and, and, uh, it's imminent for Android as well.

So it'll be across the, the full portfolio. 

Irena Andonova: Absolutely. Absolutely. We are very, very committed. So you are absolutely correct. So it is applicable to all, uh, teams rooms, uh, on Windows and on Android. Has been rolled out on both platforms. 

Tom Arbuthnot: Oh, great. 

Irena Andonova: Uh, like any AI feature, we will continue improving. It has certain limitations.

It's very, very interesting how when we develop ai, AI related features, it's kinda like, um. I, I often joke like building the plane and flying it at the same time as we have to learn. The system has to learn of how the future, uh, adopts like that. It has certain limitations that we will continue working through, uh, to make it even better for bigger meetings, um, and for more complex invitations or calendar invite.

Tom Arbuthnot: Yeah, I think it's super interesting as well with AI where we are at the moment. 'cause you can pretty much guarantee at the rate of innovation we're seeing that these models and, and, and options will get better and better. So whatever we've got today, you can imagine in 12 months will be even better than it is today doing the same thing.

Irena Andonova: Yeah, absolutely. The rate of innovation is, uh, quite staggering. 

Tom Arbuthnot: Awesome. And then like, it's not just rooms. Um, we, we had a podcast on Teams Phone a couple of weeks ago, and it was embedding AI into the, the, the phone device experience as well. 

Irena Andonova: Yeah, absolutely. Uh, we are looking at ai, uh, and delivering capabilities across all teams, devices.

Um, phones is already on the forefront. Um, can offer summarization for calls, uh, which is super useful. Uh, it's one area where, uh, we already have something in market, um, and that's. Uh, we are already hearing very positive feedback on that because again, anything that saves us time is welcomed by users.

Tom Arbuthnot: Awesome. And so that, that's some examples of kind of, uh, user benefits or business benefits. But there's a lot, and a lot of people in our community will be really interested in the, uh, the potential for operations and administration, like some of the. Teams Rooms Customers that we've had on, again, the podcast are like over a thousand rooms globally, and, and some of them are on quite tight operational teams.

Like I, I was, I'm shocked at some of the customers I talked to that run such big estates on such relatively small teams in terms of people coverage. And, uh, I, I know you've been thinking about, and there's some plans for, um, how AI can help on that side of the fence as well. 

Irena Andonova: Oh, absolutely. Operations or, um, and operations and management.

I think of it almost like lifecycle management, uh, from, uh, from the beginning towards, um, from deployment to management at scale. We think about how we can make that whole journey easier with ai, um, and also learning from the system. Uh, you're absolutely right. I hear all the time from customers. I have thousands of devices.

I, especially if it's a global multinational company like Microsoft, uh, there is a very small number of IT people who are dedicated for the rooms and they have to be able to manage and operate at a very large global scale. So, uh, and they also need to learn where the users and how the users are using these spaces so they can help the users be more productive.

So I would say we have different aspects, uh, from this lifecycle management, uh, perspective. One is, um, the pro management portal has been built from the beginning with the concept of an AI platform. In other words, the is monitoring 24/7 and it will generate an alert. If something is to go wrong, it will generate an alert.

It will generate, um, a recommendation of what action an Admin should take and ideally we'll take care of the thing. So that last part of take care of it has been a development area. It is just, like I said earlier, building the plane and flying it. That's another example. We have to continue learning these different patterns of how we can make it better, where the service can.

Ideally self remediate and get that for like example tier one intervention as close to a hundred percent as possible. So that has always been the vision for the pro management service where we can expand and not having to need human interaction. Barring any kind of physical issues, of course, but for a lot of these tier one, uh, type of issues and incidents that occur, the service should be able to take care of it.

Tom Arbuthnot: Yeah, I mean, it makes sense, doesn't it? Right. We, we are talking about all this in incredible intelligence in the AI models. If the, if the pro portal is saying remediation, recommend, uh, the reboot, it's like, well, I mean, the system knows what it's supposed to do. It knows the schedule of the room. Like, why not do it for me, waiting for me to click the buttons?

I'm, I'm really excited about the potential of that. 

Irena Andonova: Absolutely. Um, that, so that's, that's one aspect of where we'll continue innovating. Another area is we hear a lot from customers of how are the use the rooms being used, especially like even Microsoft with return office type of, uh, planning. How do we know if we have the right capacity, if rooms are used, uh.

Should they change floor plan, uh, uh, room layouts and so forth. So how do we do that? We actually started releasing, um, insights and learnings, and this is another area where we're gonna continue to, uh, help and improve the insights that we have of how do, are being used. Um, there are shared spaces, insights we can glean on and provide, like, Hey, let's say.

The rooms on the third floor are mostly used, um, on Tuesday and Thursday. Oh, in the rooms on the first floor, perhaps you are reaching a high capacity utilization. So maybe there is some need to address that in a different way for the users. So this is already rolling out in pro management portal and, um, our admins can start looking at the beginnings of how we think of these insights.

Tom Arbuthnot: Yeah, that's really interesting. And it's, um, from what I've seen of some of that stuff, it's kind of pushing beyond just like, uh, basic operational understanding into almost facilities management level of like, look at, let's look at this building. Let's look at we basic, you said there, which floors are being used, why is everybody gravitating to these rooms and they're overbooked and these rooms and they're not booked.

Like some, some higher level insights than just this room is busy or not busy. 

Irena Andonova: Yeah, absolutely. Yes. 

Tom Arbuthnot: Awesome. And like, is, is there anything else you wanna talk about or share in terms of, uh, I know we're kind of early in the next, uh, financial year in September as we record, so just getting back into the mix, but any other areas you're excited about in terms of the, the re portfolio for the coming.

Irena Andonova: So obviously we spoke a lot about AI. Um, AI is a very big focus for us, uh, from our rooms, uh, and devices portfolio. Because of the scale we are reaching, uh, we will continue focusing on, uh, helping customers deploy. With ease and manage at scale. So we will continue delivering, uh, functions like, like we have what we have enabled with autopilot on Windows, for example.

We are looking for similar capabilities on Android. Um, and then management at scale. We spoke a lot about AI, but we are also looking at how we can help, um, manage, um. Thousands of devices, even if you have a mix of platforms, mix of Windows and Android. We want to make that easy for both platforms. So that is definitely a big investment area for us.

Um. We're, um, also looking at, um, we have announced and working closely on delivering experiences on, uh, as a platform, the Microsoft, uh, device ecosystem platform, which is an Android platform, so that is a big focus for us as well. 

Tom Arbuthnot: Yeah, the direction of travel is, um, everything moving into the pro portal, isn't it?

So I think MTROW is already, uh, there. Obviously the, the Android stuff cut over relatively recently, isn't it? So that all the AI innovation is gonna apply to both the Android type and the Windows side. 

Irena Andonova: Absolutely. I mean, users, uh, our admins can already see we have start. That, um, the devices, the Android devices, panels, uh, and rooms are visible.

They're getting held signals. We all, I spoke about the AI platform, the services built on. So that will, we will look at extending the signals and expanding this platform to service across, across the devices categories. So across all devices. 

Tom Arbuthnot: Awesome. Well, Irena, thanks for coming on the show. Really interesting to hear the, the plans for the year ahead.

I'm really excited about the potential of some of that AI, AI stuff, particularly in the operations and kind of proactive, uh, management. I think that'll be a, a huge asset to a lot of organizations. 

Irena Andonova: Yes, thank you for having me. And stay tuned for even more AI and, um, other developments. 

Tom Arbuthnot: Awesome. Thanks a lot.

Irena Andonova: Thank you.