Ledge: All right. Bobby, thanks for joining us. Really cool to have you on today.
Bobby: Yeah. Thanks, Ledge. Glad to me here.
Ledge: Can you just give a quick two, three minute background of yourself and your work, just so the audience can get to know you a little bit?
Bobby: Sure. I’ve been in the industry my entire career. Even dabbled before that, so tinkered just like a lot of guys my age. We had our TI-99/4As and TRS-80s back in the ‘80s, and decided that was what we want to do with our lives.
I grew up in west Texas, in El Paso specifically. Went to Rensselaer Polytechnic Institute, so I’ve got to give them a shout out. Studied computer science there and started consulting, basically. Right out of school did a lot of consulting, which was great. It gave me an exposure to lots of different industries. Every six months got to see something brand new. New technologies and really exposure to the customer.
I would say, if I had to take, what did I get the most when I was a 22-year-old consultant? Is that there’s a difference between being in a cube and hacking away on code back then, or sitting in front of the guy who’s going to use it, whose job depends on it, definitely gave me a different perspective on why we build software and who matters to this whole equation.
From there it’s the traditional. Went through the dot com boom and bust in spectacular fashion. Got to see the entire thing do its cycle there in Silicon Valley at another consulting company. After that, it’s mixed in.
Then I got afraid of enterprise, did a consumer product and that’s where I learned, that’s where I found Python. That’s my weapon of choice to this day – as a matter of fact to this hour, probably. Right we get on the phone here. From there it’s been a lot of different things.
A touch more consulting. I did security for a little while – did a CMS thing. Then, four years ago, joined Lifesize. You may not be familiar with that because we’re using Zoom to do this call, but I’m not going to hold that against you. I’ve been the CTO here for a couple of years.
The interesting thing about video is that I’ve never done video before. It’s a very different world. Traditionally, you’re okay if something takes a second. If you go to a web page and you hit Okay, and you hit a Submit and it comes back in about a second, that’s alright. You can’t do that in video. You have about 33 milliseconds to actually do anything interesting, and it’s a whole new set of problems.
Trying to have thousands of calls going on at one time. Trying to do billions of minutes a year of calls. Knowing that you can ruin someone’s meeting if this call goes poorly. We’ve set up this time four weeks ago, and if something happened to the service or something like that then we’d lose 30 minutes and it’s a complete waste of our time. So it’s imperative that we keep this thing running.
It’s very different. A very different set of problems. We’ve been doing a lot of really cool things here. It’s been a journey of reinventing the whole technology stack. We’ve actually gone down the path of paying off our technology debt, and I think we can safely say we’ve paid it down to where we can manage it now. In the last two years, every line of code that ran the Lifesize platform has been replaced in some form. Our servers have been replaced, our clients, we’ve got a brand new set of devices that just came out a few weeks ago. It’s all 4K video in the cloud. New hardware, new software, we’ve done it all in the last two years. It’s been pretty cool.
That’s my deal.
Ledge: We are huge advocates of the remote work paradigm, and this is just a critical tool. So, yes, we’re using a competitor tool here. I’m going to go cower in shame.
Bobby: That’s okay.
Ledge: Tell me about, how do you guys do this? I’m a relative technologist and was a software developer way back when, but I find it extraordinary the way that these tools have evolved. Certain other competitors – people will send us links for, I don’t know, a Google Chat or Skype or what have you, and the new breed is so much better.
What is behind the hood? These things are incredible. Just thinking about real time 4K video over a relatively decent connection, how does that even happen? Maybe walk through the high level of that technology stack.
Bobby: Sure. We’ve fully gone all in on WebRTC. Google kicked that off. Inside of Chrome, you’ll find WebRTC so all the browsers are starting to support it.
We took a standards based approach. Everyone had their own proprietary stack before. As we moved over to a new stack we decided that being open and not closed is really the right way to go. What that helps us with is, everyone’s working on this problem. When you make a call, it goes over to the stack. Then the interesting part – what I would call the secret sauce of being a provider – is not just getting the bits across, because ultimately what you’re doing is I’m taking a bit stream and it’s up to me to decide how I’m going to deliver it to all the participants on the call.
Once you solve some of the basic problems of moving media across, which is stuff that we’ve been doing for a long time, the real hard part here is just joining the meeting. Everyone worries, how do you get 50 people to call at once and talk and to do that?
If you’ve ever been in a room with 50 people that are all talking and try to manage that? It’s worse when you’re on video, or even worse on a phone call.
Yeah, it’s pretty cool what we do. We certainly have done a lot of work on interesting video technology and how can you rapidly mix. If you think about a 30 frame-per-second call, you divide that into how much time do you have? You have not very much time. You have milliseconds to take a frame, mix it with everybody that’s on there, and shoot it back out to every individual. Or, depending on your architecture, traffic coping all the bits and sending them everywhere.
That part is much easier because you have powerful clients today. I can run my entire stack inside of Chrome, which is great. You have the browsers coming along, and so the technology is there.
I see the big problem as not so much the… The quality of video is getting very good, we’ve got 4K, but joining that meeting. The experience right now, we think it’s something like 6 to 7% of online meetings done with video. That means there’s still, in 2019, 93% of people are still deciding to pick up the phone and talk to each other in audio conference when we have ridiculously amazing technology across all the competitors.
Some of the things we want to do is, we’re trying to figure out how to make it easier to join. It’s interesting, when we talk about, how do you call someone on video? It’s five steps today, often. It’s like, “Okay. Well, do you want to do this on video?” “Yeah. I can do this on video.” “Okay. Let me get this calendar invite together.” “Okay, I’ll look for that.” “Alright. Download this software.” “Okay. Let me do that.”
It’s work. It’s too much work. I’m just going to pick up the phone and call you if I don’t know how to contact you on video.
I think what I worry about, the next of wave of stuff we need to solve is how you find each other. You can choose to use Zoom or BlueJeans or whatever and Teams is a big thing – Microsoft Teams are pushing that thing pretty hard – we have to be able to discover each other. I don’t ask you what cell phone provider you use. I don’t ask you where’s your email hosted. I don’t care. I just have a simple way to contact you and it just works.
So far, we’ve spent the last ten years making this call work so that this part looks good, but now we have to spend the next few years figuring out how to actually connect people this way so that it becomes natural. When I talk about what’s hard on our business, the video part’s actually been figured out. We’ve been doing this since 2003. It’s not hard to move bits around a wire. It’s hard to get people to actually feel comfortable using it and to stop saying, “Can you hear me?” “Are you there?” “Can you see my screen?” Those are the things we just want to get rid of and have it easier to connect people.
Ledge: Talk about the R&D on that. That’s a paradigm that certainly not just connectivity and communication providers are going to want. I want everybody to access whatever I’m doing extremely quickly, and I want to know who they are and be able to meaningfully provide that user experience.
What’s that like? That R&D sounds amazing.
Bobby: How do you simplify the experience, is really what it comes down to. This requires some industry support, but basically all of us have built these silos. We claim interoperability, but that just means that our devices can talk to each other. But really, if you’re going to use Zoom and I’m going to be on a laptop, I have to use Zoom to talk to you. Same way with Lifesize. It’s not like a Zoom person can just call Lifesize from their laptop. No. They can call my conference room, but certainly not each other.
Discoverability is really the hard part. How did we solve this problem before with the phones, or with email addresses, and why do I need anything more than that, actually? If you walk up to a random person, I don’t care if you’re an iPhone or Android user, but if you talk to someone that’s not – if you’re an iOS user and you talk to an Android, how do you do a video call with them? What do you do? What would be your steps? You would be like, “Should we use Zoom or something?” There’s no intuitive way to do it. If you’re both on iOS, fine, you can FaceTime for point to point.
I think we need to use the bits that are there because everyone has a way to contact each other. You have my phone number. You’d have to text me if you wanted to get a hold of me, and maybe we could do a video call. I think what we need is, the industry has to step up a little bit. We have to be a little bit more open.
Lifesize has to do it too. We have to say, listen, if you’re trying to reach someone then you should just look for them. I just want to look for B Beckmann at Lifesize.com. I don’t want to type in an extension. I don’t want to type in a… I want to type in a 10-digit number and start this meeting. What other industry do you have to still remember numbers? There’s no one. Do you know anyone’s phone number? I barely know mine and maybe my kids’.
But here we are, “No, no, no. If you just type in this 10-digit number at this app someone will magically connect you.” We have to just get beyond that. Part of it is, it’s less R&D, it’s more understanding, how can we get everyone on board to do things, to work together. To not worry about that.
If we’re talking about a 6% penetration rate on video, who cares about competition at this point? It should be, let’s give the customers what we think they want and let’s all find a way to work together to get that number… If we get that number up to 20%, we’ve all tripled. That’s fine. If our business triples, who’s going to complain about that? If everyone does, I don’t think we’ve got a problem.
I think those are the problems to solve. Is how do you get everyone to stop building these walled gardens around video which are kind of the last bastion of in the internet and play nice with each other.
Ledge: Well, and then this argument of, all their net services are centralized-decentralize, walled garden-open. You’re getting down to the root, classical, individual problem, and anonymization and all these things. This is ultimately where blockchain identities are going to be interesting.
How are you going to propose to solve and get everybody else on board? What’s the best-of-breed idea in this space even?
Bobby: I don’t know the best-of-breed idea yet, but what I’m going to do is I’m going to start opening it up by letting you just connect with people on Lifesize. So instead of of us setting up a meeting, then you’d be able to just imagine, go to our website and type in a phone number. That phone number automatically looks for you.
Does it know that are you a Lifesize user already? Great. I’ll just ring your phone. You’re not a Lifesize user? Okay. Then I’m just going to send you a text with a link and you click on it right from your browser, the call starts. You don’t download software. You don’t do anything. You have an amazingly powerful computer in your pocket.
But type in an email address. Great. I send a link to you in your email. We try to get people to just start using it that way.
We’re going to just start doing that. Will it be the best? I don’t know. Every one that comes out, there’s no best yet. I think the idea is to just give it to people, try some things, and then see if it works.
We’re talking about, just come to our website and try something. We’re not telling you to sign up for an account. We’re not telling you to do all these things, just send me an email. You wouldn’t even think twice about that. Well, great. Just make a call, don’t think twice about that. That’s what we’re going to do.
Then we’re going to say, hey, why not open up to these open platforms that we have? Why can’t they communicate? We’re all big proponents of open source here. We use WebRTC. Our clients are fully open source. We’d love that fact that other people would want to talk to us.
Ultimately, it requires customers to say, I really want this. I can come up with the shiniest ball in the group there, but I need to make sure that the people want it. So I’m doing the whole, oh, I don’t know if they want it but I’m going to give it to them and I’m going to hope that they start using it, and I’m going to see what happens.
Ledge: It’s Steve Jobsian – users don’t know what they want, so let’s give it to them anyway.
Bobby: That’s right.
Ledge: That’s excellent. I can speak from my own experience in sales, services and the podcast work, that the video connection is critical to what we do. I can say, without question, that I develop and maintain major accounts and relationships with people that I’ve never been in the room with. I encourage everyone to really get on the video train because I think it makes a tremendous difference. I ask everyone to do that for every call I’m on because I do believe that, if we’re going to embrace remote work and we’re really going to behave as if we sort of are colleagues and not with a line of separation or this demarcation between locations, that’s the way to do it.
I’ve even talked to people who are using video solutions for hybrid teams where they actually will set up workstations that are entirely always on, connections to whoever’s at home. I wonder, do you have clients that work that way, because that seems to be the next generation where there’ll be this ubiquitous video terminal to all people at all locations.
Potentially then, talking over that connection in a one-hop kind of way where I would look over to the next cube and talk, except the person wouldn’t be the person, the person would be the video.
Ledge: It seems like audio is the biggest issue there in the environmental.
Bobby: A lot of noise. Yeah. We have customers do that today. We have a hardware solution and they’ll have a full-time camera running in their lunchroom, for example, so people can see each other in remote cities.
We certainly have team rooms. Even at Lifesize we have an office in Raleigh, North Carolina. We have one of our devices right in the middle of the team room. We can fire it up any time and just have a big group discussion.
Noise is obviously an issue if you start talking and there’s all this background noise. But there’s a lot of interesting work in, how do you tune that out of the background noise. There’s these neural networks out there that have blown away the echo-cancelling in the last ten years. New neural networks that have come out. Mozilla has a really nice package they’ve open sourced and it does a good job of getting rid of background noise. So you’ve got people chattering away in the lunchroom and you’re maybe trying to have a conversation with someone on video, it won’t overwhelm the conversation. Just like in real life, you can focus on each other.
Interestingly enough, we’ve moved into the higher resolution. When you start to talk to people over 4K on large screens, the reality, there’s some sort of strange – I don’t know if it’s some sort of biological, I have no idea, and it’s not my field at all – but I can just tell it’s a different conversation. As these resolutions get better, as it feels smoother, it feels less like having a video call as opposed to talking to you through a window. I think those things help.
I think the big blocker to a lot of this stuff is privacy. I know who’s in the room, but are the cameras watching me? Where’s this going? I certainly am not interested in always watching people in doing things, so I think solving those problems, it’s some of the interesting things that a lot of us do as vendors. Our camera turns away when it’s off, so you know it’s off. It’s looking away and down. But how are you sure that the camera’s not listening to you or doing that?
I think people, will they be themselves? If you and I are just sitting casually in a room talking, and we know there’s a video camera that’s constantly pointing at us, what if someone’s off-frame listening?
I think there’s a lot of things about that side of the world that we still don’t have a good solution for. That’s where we’ll just have to see what happens and how the industry progresses, and what we can do there to ensure that people can feel safe. If you feel like you’re being watched all the time, you will change what you say – and I think that’s not good either.
Ledge: It seems to me there may be IoT sensor, sort of ubiquitous detections mechanisms and such. Where it’s not just about is somebody in the frame or is somebody on, but have the lights turned on in the office therefore there’s somebody there? Or, connected devices that it really becomes more ambient and environmental, and there are signals coming in the way that would not be… You could connect; did the door open? Are chairs moving around? Is there a butt in the seat?
Bobby: Yeah. We’ve done stuff like that. We’ve done some basically low-res wide angled cameras where you can’t really make out who’s in the room but you can tell there’s motion in the room. That will sit there so they can tell the camera to do something different. So that you know something’s going on, but maybe not have enough data to ascertain who or what so that you still have some measure of privacy.
But these things, we sat around here I remember a year ago talking about machine vision. AI was the big theme in 2018. It was like, how fast is machine vision progressing? Even I was like, “Well, it’s moving. You would be blown away with what it is in one year.” Within four or five months, I can go to Target and Google ‘sell me something for 40 bucks’ it will tell me the. It was moving, it’s jumping in terms of months at this point. I think the conversation we have today, we’ll look back on it two or three years and be like, “Do you remember when we were talking about, how could that ever possibly work?” I just think that, in those areas there’s a lot of moving forward.
Ledge: How about AR/VR? Are there discussions around those? I know it was the hot, hot thing at CES a couple of years ago and now it’s kind or wonk wonk.
AR seems really interesting for meeting and collaboration. That I might be able to grab an arrow and move it around in my video or something.
Is there any discussion around that?
Bobby: There’s some of that, we just don’t get a lot of customers asking for it. The interesting thing is that they’re still asking for some basics. Green screen is a kind of cool thing that you pick out there, especially when you get to some of the higher res stuff you can do some neat things with green screen.
The AR thing I think will be interesting in some certain specialties. There’s plenty of people that are out there experimenting with it, but commercializing it I think there’s probably going to be some specific industries where it makes a ton of sense. Where you have these super high resolution cameras. You’re able to then maybe put on some sort of source, then walk around and that person can really see what it is that you’re trying to do.
You can imagine remote repair, or you want to go to the top of one of those air windmills we have all over Texas, but I’m sure somebody has to. Even if they have a problem and they just clip on some lenses and all of a sudden in 4K I get to see what he’s seeing. There’s all sorts of interesting things that you can imagine where you can then show him what you need fixed.
I think those are coming. I think those are the interesting downstream things. We still have a lot of problems with the conference room. We take that technology and then we find out where it makes sense.
Today, we’ve got people that still want to do something as simple as, I just want to be able to see my whiteboard in the conference room during the presentation without having to turn cameras all the time. Stuff like that. There’s some cool camera things doing some neat things, where they are doing some image manipulation. You see that little painting there behind you? If that were a whiteboard, that camera would actually be able to go find that, clean it up and present it as your presentation source without you having to go and muck around with your cameras and try to clean it up.
There’s neat things like that which you think, wow, we’re still trying to figure out how to share whiteboards? But yeah, it’s still pretty hard to do that. But we’re coming along. A lot of these things are being worked on and so there’s a lot of new things coming out from all these ventures.
Ledge: Are you guys heavily invested in using public cloud? Is that how you’re hosting all these things? Anything special there?
Bobby: Yeah. Amazon is our horse. We do a lot of things in AWS. That’s where we’ve rolled it out. We’ve just replatformed everything. We’re actively working on changing it all up, and we’ve moved it all to AWS.
The big concern with any media running on public cloud is the amount of latency and you’re sharing CPUs and stuff like that. Those have mostly gone away. We seem to get very consistent behavior out of the cloud, and we’re able to move it out there. It reduces so much overhead and makes us move so much faster.
Amazon rolls out so many services pretty much every year. If you ever go to reinvent, it’s just overwhelming. We use it all. I look at the stack of things on my bill from Amazon – it’s everything from the instances all the way up to very interesting things that our engineering team is playing with – and really taking advantage of those things is making us move faster.
To replatform something like what we do, and to do it all in roughly 18 to 24 months is really hard to do. By replatform I mean we ripped it all out. It’s going to be all gone. It’s not easy and you’ve got to really be careful because you can’t really interrupt services. It’s a lot easier to start a new company than it is to say, hey, thousands of you, we’re about to just change your show over tonight, and hope it works.
We’re all about using Amazon, and it’s a very great partner of ours.
Ledge: How did you mix the technical debt, investment, remediation, with the need to develop new features and keep up with the product roadmap? You can’t just halt progress for 18 months and two years while you redo everything.
How did you balance that, from a resource perspective and from just a product roadmap, even down to scrum teams or whatever you were doing?
Bobby: The way we approached it, we had two phases. One it was going to rip out the clients and replace them with brand new ones. That was the first phase, that finished about a year ago. The pitch there to sales and company was, this will let us move faster when it’s done but we have to take a six-month break. You’re not going to get any new features for a little while. You’ve got to sell what’s on the truck, and I know that it’s a hard conversation to have but at the same time it’s important to show why you’re doing this and what you get on the backend, and actually delivered.
The flipside of that is, we delivered the new clients and we showed, look how much faster we can move now. It paid off. Now we’re going to do the same thing with the server.
Now, changing out the servers meant we still had some leeway. We have a lot of features, a lot of stuff on the edge that’s not necessarily part of the prime video experience. We have wireless data sharing, some mobile stuff. We had enough other things. We came out with a new kit.
What we did was, we just built things on the periphery that still made our product line fresh and brand new and we came out with these 4K devices. So while we don’t have to mess with that core video experience for a little while, and we can roll these other features out and then rebuild this. These new products and these new features keep the sales team busy for this year. Then, as we roll this new one out, now we can shift back over and start adding features to our core server platform.
Half of it was a negotiation. There was a time when we did say, six months no new features. Then the other half was like, okay, what can we build when we know we can’t touch this?
An interesting story about that is, at the beginning of 2018, in January, we didn’t know what we were going to release because we were going to spend all of our time building this new platform called Galaxy. So what we did, and we pitched this to the company, is, guess what, engineering is taking two weeks off. Meaning we’re not going to work on anything that has to do with any shippable code. We’re going to take two weeks off and we’re going to have what we call an innovation sprint.
We let anyone pitch an idea, and if they could get a team to join them they could spend two weeks working on whatever they want. Out of that came our entire 2018 releases. Pretty much everything that we released came out of that two-week break of just put everything away and let’s just see what comes out of all these people who have all these idea.
The biggest users of our products are our employees, and they all wanted things. They pitched great things and out of that we have something called Dash, which is our new solution that came out of that. Lifesize Share. All sorts of really neat things. Some cool voice control stuff.
Just out of doing a two-week break and shut it all down for two weeks and let the people do what they want, paid off. The dividends were huge. The ROI was amazing on that. I think we’re going to do another one here in a couple of months once this phase is done.
Ledge: That’s fantastic. To have that level of buy-in across the organization for such a thing, that just sounds radical. What? You’re going to stop? Engineering is always behind. We can’t possibly move fast enough.
How did you pull that off politically? You must be a good diplomat.
Bobby: Well, I think we had had a successful release of the clients earlier, when we had said that we were going to take six months and then new clients getting released. It was the start of the year. We didn’t have a whole lot to announce – so that was the other thing. Like, well, we need to do something. I think the other part was…
We had detractors. There were certainly people like, “Wait, engineering is not going to do any of my stuff for two weeks?” I’m like, it’s literally two weeks. Two weeks is not a long time. We’re going to be okay. If we stop working for two weeks and this whole company comes crashing down, something else is wrong besides engineering.
So we made that pitch and we said that, you can be involved too. There’s nothing to stop you, Mr. Detractor, from showing up to this day. It didn’t have to be an engineer that kicked this off. Anyone can pitch an idea, and if you can find five to seven people that agree with you that this is an idea worth investing their two weeks in, then they’ll go build it. You can do whatever you want.
That pretty much shut that down, because they’ve got some great ideas and, come on down and pitch it but don’t just sit there and say this is not going to work because that’s not going to fly. We have to try new things because we’re not going to move the company forward.
It paid off. It paid off not just in the ROI for the business, but our attrition is probably a third of what it is at other companies in Austin. I think we’re like a sub-3% in engineering in 2018. Which is hard to do when you’re in Austin and you have the Amazons and Facebooks storming your gates, throwing out their big money.
The fact that people want to do this and work here and do that, it says that this is a good approach to how we do things here. In cultural it’s taken hold, and I think now proving that it didn’t matter who you were in the company, if you could pitch something and it took hold and became something then it means that anyone should do that. You don’t have to wait for an innovations sprint to innovate.
I think pushing that culture across the board, top to bottom, it doesn’t matter what part of the company you’re in, is something that we pushed and it’s taken hold. It’s for real. I know people talk about this stuff all the time, but we can actually show it and measure it and have seen something come out of it on the other end.
Ledge: That’s excellent. I love that. For the culture and the leadership it’s fantastic. You don’t hear it very often, it’s the stuff you read about in articles. So, well done.
I’m curious, as a business systems thinker across the enterprise, have you had any other groups in the company – the operations groups, HR, finance, anybody like that – say, hey, we want to do that. We can do that for our area of the company and remove process. Has any of that taken hold?
Bobby: I think a little bit in our IT systems. They restaffed and retrained and did some different things there to try to break down their problems and be a little bit more innovative. Any large software company like ours has gone through a transformation from pure hardware, to a SaaS, to all those things. Our business systems graph is scary. If you look and see what runs the whole business it’s like, wow, you need all those things?
When you look at that – and just like in the– when you look at it, it looks like it’s way too big to do anything about. What they did is, okay, let’s pick a square. Pick a square and let’s attack that square. Instead of being spread too thin, let’s just do what we did with the service thing – we’re not changing that for the next six, but over those other squares, they’re kind of the way they are. We’re going to have to get our processes to hold together and use enough duct tape to keep things rolling, but we’re going to over here and fix this square. We’ll see you in a few months. When this square is out you’re going to be so happy that you’ll be okay with what you had to do in the last six months, and then we’ll go move on to another square.
I think, at a cultural level, it does seem to stick and I think we’ve seen it. We haven’t had a whole department take two weeks off like we did, I think that was a little aggressive on the engineering side, but I think the idea has stuck and the mentality is there.
Ledge: That’s outstanding. One of my favorite questions is, so if I gave you a piece of paper and a Sharpie and say, you can write a letter to yourself 10 years ago. What are you going to write on that?
Bobby: Let’s see. Ten year ago. 2008 – no, 2009. It’s a great question. Let’s see. 2009. What was happening in 2009? The recession ended. World was ending.
The first thing I would say is, everything that you’ve done so far has been little compared to what’s about to happen over the next ten years. Don’t get hung up on the ups and downs, the company didn’t do so well, or any of the other successes in the past because we’re about to see a whole new sea change in what’s going to happen. Buckle up and keep sharp, because there’s a lot of really cool things coming. Don’t stop writing code – which I still do every day. I didn’t fall into the trap. Something along those lines.
Bobby: And the Cowboys will never win the Super Bowl again. Just to get it out of the way, that way. That way I wouldn’t be so sad every year. That’s probably the other thing I’d tell myself.
Ledge: I’m a Jets fan, I feel you.
Bobby, great having you in here. I love the insights. Thanks so much for joining us.
Bobby: Yeah. Thank you.