AIMinds #036 | Bruce Sharpe, Chief Product Officer at Humach

About this episode

Bruce is a seasoned innovator and technology leader with roots in the enterprise and startup environments.  Bruce has been able to lead and drive technologists from application ideation and development to operational deployment in cloud-based environments.  

Leveraging his engineering background, his agility, experience, attention to detail and leadership has allowed him to hold various senior positions in a number of companies and industries.  

Bruce has earned a reputation for consistently delivering results across multiple industry sectors including Technology, Financial, Hospitality, and Healthcare. His career has included positions of technical leadership at Ticketmaster, USA eCommerce & Services, TTEC Holdings and currently as Chief Product Officer at Humach.

Bruce is passionate about innovation and change for today’s CX environment.

Listen to the episode on Spotify, Apple Podcast, Podcast addicts, Castbox. You can also watch this episode on YouTube.

In episode of AI Minds podcast, hosted by Demetrios, Bruce Sharpe, Chief Product Officer at Humach, delves into his extensive experience in telecommunications, AI, and customer service innovations. From his early days experimenting with electronics, influenced by his father, to leading innovation and becoming a patent expert at Teletech for 18 years, Bruce’s journey has been marked by continuous technological advancement.

Bruce discusses his critical work in developing omni-channel platforms at Ticketmaster, managing telecommunications for large-scale events, and patent processes involving efficient collaboration with giants like Microsoft and IBM. Transitioning to AI, Bruce emphasizes the importance of voice interactions, partnering with Deepgram to enhance digital agent capabilities in providing business support.

The podcast also touches on the challenges of deploying AI in regulated spaces like healthcare and insurance, where Bruce highlights the necessity of rigorous testing and the indispensable role of live agents in managing sensitive inquiries.

This conversation gives insights into a blend of human and machine interactions in customer experience, revealing practical approaches to integration and the critical evaluation process for technological partnerships.

Fun Fact: A compelling anecdote from the podcast was when Bruce Sharpe, despite clearing an interview and security checks for a job at Martin Marietta Aerospace, was dismissed on his first day because he was not yet 18. He was just one month shy of the age requirement!

Show Notes:

00:00 Exploring future companies. Sponsored by Deepgram.

03:13 Young guy excited to prototype missile guidance.

08:50 My contact center career began with Redline.

12:49 Unforeseen challenges eventually led to Teletech job.

14:51 Virtual expert retail assistant kiosk in 2006.

18:40 Turn ideas into detailed, documented patent applications.

22:09 Deepgram aids our digital agent strategy effectively.

23:35 AI supports, live agents handle complex or unwilling customers.

28:35 Challenges in evaluating and integrating partner technologies.

31:50 Cost, longevity, achievements, funding, methodology, efficiency, reliability.

34:21 Thanks a lot, this was awesome.

More Quotes from [Speaker]:

Transcript:

Demetrios:

Welcome back, everyone, to the AI Minds podcast. This is a podcast where we explore the companies of tomorrow being built. AI first, I am your host, Demetrios, and this episode is brought to you by the number one voice API on the Internet today, Deepgram. It is trusted by the world's top enterprises, conversational AI experts and leaders and startups, including but not limited to, companies such as Spotify, Twilio, NASA, and Citibank. Today on this episode, we are joined by Bruce, the CPO, or chief product officer, as some people like to call it at Humach. How you doing today, Bruce?

Bruce Sharpe:

I'm doing well, Demetrios, very well.

Demetrios:

I'm just glad that I was able to say Humach correctly. I really appreciate the two words being smashed together and you telling me that it's human and machines.

Bruce Sharpe:

We're on the forefront with all the AI that's going around. We're on the forefront of the humans and machines. Never want to leave the human out, right?

Demetrios:

That's it. That is exactly right. So you've got some cool stories that I want to get into before we talk about humach, one of which is your stint in aerospace. Can you tell us about that?

Bruce Sharpe:

That's a pretty funny story. Yeah. While I was in high school, I actually was able to get into electronics. Started my first day, actually. I went to my dad and said, hey, they've got this course that they want me to take as a shop type of class. And the course was basic electricity. And I told him, is that a fair trade to actually get into? And he's like, yeah, thats a really good trade. You ought to check it out.

Bruce Sharpe:

So I went in, I applied, I got to the class the first day, found out it was basic electronics, and I was flabbergasted. I went back to him and go, I don't know anything about this. What do you. And he's like, I don't know. Give it a shot. So I did. Anyway, long story short, fast forward four years in high school. I actually graduated with an associates for electrical engineering.

Bruce Sharpe:

And what I did is I applied for a job, a design engineer's position at Mark Marietta Aerospace in Colorado. So, and I'm from Denver. So anyway, I went through this long process of, you know, clearance, activity, evaluation, stuff like that just to get in to do the interview. So they chose me. They did the background check, took about five or six months. After that, I got in, I went and interviewed. They walked me around, checked everything out, and they dropped me off in a room. There were like eight people in this room.

Bruce Sharpe:

It was like a panel of judges, but what they were doing is just firing questions, right? And I apparently did pretty well because they told me at the end of the day that I was hired for the position. So I thought, wow, this is awesome. I'm a young guy, I'm getting into aerospace. It was my dream where I wanted to go, and I go home, I tell my folks, I'm so excited. I go ahead and get set the next week to start my job. I go in the first day, the first hour, they're walking me around, checking things out, showing me the lay of the land, seeing these giant cylinders that they put satellites in to do shake and environmental testing on. Anyway, get all the way over to meet with one of the engineers, and he's got a design and he wants me to go prototype it. Well, what I didn't know at the time is that I was going to be prototyping infrared guidance for MX missiles.

Bruce Sharpe:

So I started doing some work and I met with him for an hour. I went in and started doing work for a little bit, did lunch, all the standard things that you normally do your first day, HR, etc. But come about hour seven, a couple people walked into the room. It was the HR person who had given me the news that I was hired, and it was a security guard. And I'm like going, whoa, what's going on? They walk over and they say, Mr. Sharpe, unfortunately we've got a problem and you're going to have to come with us. So I went with them. They got me into a room and said, as you know, this is a government project. It's going to be staffed for four years.

Bruce Sharpe:

We have to ensure that the staffing is set by a certain date, which is like a week away. So we thought we had all the positions filled, but we've got a problem with yours. And that is since this is a government job, you have to be 18 years old. And I'm like, what?

Demetrios:

You were at 18?

Bruce Sharpe:

I was 17 years. Eleven months.

Demetrios:

No.

Bruce Sharpe:

And why they didn't catch that in my security checks and everything, I have no clue. But anyway, they said they're going to have to let me go, that they can't staff me legally because of my age. And I'm like, you mean you can't just wait a month and I'll turn 18? And everything is like, no, because we have to lock it in in the next week and it's staffed and we've got a tight schedule. We can't delay anything. So anyway, that was my one day in aerospace. I actually got into the lab, I started working, and then I got my legs cut off at the kneecap saying, I'm sorry, you're too young. But.

Demetrios:

So one thing that I often love to think about is how moments like those in life, in the moment, it's very difficult as we go through it. I'm sure you had your hopes up, you had a lot of expectations around that, and it came crashing down in the span of probably less than an hour. What were you able to do after that? Because you were free and you weren't in a job.

Bruce Sharpe:

I was devastated, number one. I mean. Cause it was the first time I had actually been, let's say I was laid off, right? Yeah. So that was the first time I immediately had to go in and start hitting, you know, the job boards and things of that nature to and even try to resurrect some of the positions that I had tried applying to because I wasn't sure that I was even going to get an opportunity to get into this position. So, yeah, so that it was challenging. It was like starting all over again to like.

Demetrios:

There was silver lining that was in that though.

Bruce Sharpe:

Yeah. In some ways, even though I would have loved to have been in aerospace, I mean, like I said, it was a dream of mine. I knew I'd never be an astronaut because of the, you know, glasses and all these other issues that you have to be perfect. I did get into a really interesting field. I got into telecommunications, so I did get to do hardware engineering for about ten years until I started having power system components explode in my hands. And I'm like, I'm out of this, I'm done.

Demetrios:

So I ended up, wait, what's the story there? Why were they exposed?

Bruce Sharpe:

So in the startup company, it was called Chernett Communications. I had actually started working on what's called answer supervision. So at the time, a lot of the phone systems didn't have positive answer detection. So you'd go to a hotel, for instance, a business, and you would make a phone call out of the hotel and then it'd be 1 minute long, and you'd go to check out and you'd find out you've got a five dollar bill for a phone call. And you're like, what? Well, they use this thing called at and t operator assisted rates, which was a rating table for calls, and it was based on the length of the call. It doesn't matter whether it was answered or not, you were charged for it. So we developed a technology that allows us to, you know, based on an algorithm of cadence, different tones, a lot of things like that. We were able to identify positively when somebody was answering a call or not.

Bruce Sharpe:

That even, and this was back in, geez, like the, the late two thousands or excuse me, at the late 1990s, somewhere in there.

Demetrios:

Yeah.

Bruce Sharpe:

But uh, we were able to identify things pretty quickly that it was a live human or an answering machine, whatever. But anyway, long story short, um, that kind of led me into the contact center space because when I left there I ended up going to work for a couple small companies, but ultimately ended up at redline hotels doing telecommunications. I was working on the technology and supporting it as opposed to developing it. So I got really good at that. And they decided that they were going to open up a central reservation network and they needed somebody to come in and get trained up on the software and the technology to be able to implement and set, you know, for their 80 or so properties. Anyway, I got trained in aspect communications gear and I became a certified engineer on the contact center. So a phone system and that kind of led into a lot of the other things that I've done over my career. So I mean, I guess in a way there was a silver lining.

Bruce Sharpe:

It kind of led me to where I'm at. Wasn't where I thought I'd be, but I'm okay with it.

Demetrios:

When did you start picking up patents? Because oh jeez, you've got over 47 of them.

Bruce Sharpe:

Yeah. It wasn't until say, when I left Red Lion I was picked up by a Ticketmaster and Ticketmaster was looking for somebody to, to expand their network. They had just signed an agreement with aspect. Funny thing, I met their COO at an aspect conference. He bumped into me and he was out there recruiting. He wasn't trying to learn anything. Took six months, but he finally convinced me to come on board.

Demetrios:

So it was worth it.

Bruce Sharpe:

Yeah. So I ended up doing a couple years of technology. Then I did the, I was there four years. The other two years I did a lot of things around omni channel platform. So being able to support in a contact center, you know, email chat, web traffic, voice calls, you name it, inbound, outbound. So I got really good at that.

Demetrios:

And I know it's for Ticketmaster that.

Bruce Sharpe:

Was for Ticketmaster that was left or.

Demetrios:

They had their support be tested by the Taylor Swift debacle. Rent.

Bruce Sharpe:

I could go, actually. Yeah, it was before that actually. There's some really big events that happened way before Taylor Swift.

Demetrios:

Like it's not a new thing.

Bruce Sharpe:

I could tell you a whole story. No, I could tell you a big story about Garth Brooks that I ran. Oh yeah, Dallas, Texas at the time. Yeah, it was pretty crazy. So we were actually running a quick story. We were running a, what they call an on sale, and it was to sell a number of tickets for a specific artist. Well, we found out a couple of weeks before it actually went on sale that it was going to be Garth Brooks. And at the time it was a return that he was making.

Bruce Sharpe:

Well, long story short, you have to do a lot of preparation with the telecommunications, with all the different carriers. I mean, I had like seven or eight different carriers, local and long distance, that I was having to deal with. And because people were calling in, yeah, they, they would call. The tickets would go on sale like at 10:00 a.m. on a Saturday morning, and people would call and they'd just blast these phones. Well, one local area couldn't take all that traffic, so we actually had to take it, forward it to 800 numbers, and then drop them off at a bunch of different contact centers across the United States to actually handle. Because it was so volume intensive. We were getting like every 15 minutes, we were getting 15 million attempts, callers calling in.

Bruce Sharpe:

It was massive. Anyway, when I started the job, I hadn't appreciated the amount of work that went into that. And what I had found out is a year prior with another, uh, artist, that they had actually shut down one of what they call the five e switches. Its the big central offices that run a big metropolitan area, and it, it locked up the system so that even 911 calls couldn't go through it. So it was, it was a very ugly, very ugly situation. But long story short, I came out of it unscathed, but a couple other technologists that ran different parts of the country didn't do so well. So anyway, that really led me into working for Teletech. So I went in there, they found out a lot of the stuff that I had done in my early days around multi or omni channel contact centers, being able to do routing of telephone numbers, Internet networking switches.

Bruce Sharpe:

So they brought me in and I worked with them on a startup, and it went well for a few years, and then they decided to actually bring it into the business and actually grow it bigger. So I ended up running innovation for Teletech. I was there about 18 years. I reported to the CTO and dotted line to our CEO at the time, actually he still. But yeah, about five or six years into the work that I had done at Teletech is where I got into patents. So with a lot of the innovation that we were working in and the vendors, the companies that we did work for. So there's companies like Sprint and Walmart and Best Buy and Bank of America. A lot of these big companies we were doing work for at the time, and I was working on different technologies.

Bruce Sharpe:

Like one project that we called codenamed Vera. It was a virtual expert retail assistant. And it was like you going into a best buy and not being able to have an individual to help you out with a purchase, you didn't know where to go. So what we would do is we'd set up these kiosks in the store, and you knew that you could go over, you could use the kiosk to poke and prod and try to find what you're looking for. And with some logic that we put on it, after a certain amount of poking and prodding, you would be prompted to get virtual help. Well, as soon as you clicked yes, boom, it popped up with a video, and you're online with a remote expert, and you're talking to them and doing things. So this was back in the. We were doing that in 2006, 2007, before a lot of these other things were ever done before.

Demetrios:

Zoom.

Bruce Sharpe:

Yeah, Zoom wasn't even lower in anybody's eye at the time.

Demetrios:

Wow.

Bruce Sharpe:

Yeah. So that was one of the technologies that we actually built with a few partners, Microsoft and IBM, etcetera. But we did a few patents on that, on that capability. But that's when I really got into the patents is like 2005, 2006, and that one went well enough that we started doing more. So I got into RFID technology for positioning in a store. I got into the virtual expert. We built a couple products internally that we then patented, etcetera. So I got into a lot of different areas.

Bruce Sharpe:

I ended up becoming the patent authority for the company. I was the exclusive guide to bring any ideas to hand. Spent a little bit.

Demetrios:

You would help someone get a patent.

Bruce Sharpe:

Yeah.

Demetrios:

Then on that virtual assistant, was it happening through the phone lines or was it already on the Internet? It wasn't Wi Fi. I can imagine. It was dialogue.

Bruce Sharpe:

It wasn't on the Internet. We actually had direct connectivity, like t one or I circuits that we'd actually have going into each of the facilities, and then we'd run a private network. So we didn't have to worry about latency or anything like that, because that latency was a bigger problem then on the Internet than it is now.

Demetrios:

Yeah, that's why I was wondering, because it feels like that's a really hard nut to crack.

Bruce Sharpe:

Yeah, definitely.

Demetrios:

And when you say t one. I remember those days especially because for anybody that did not live through those days, t one was so fast. When it first came out, it felt like, oh my God, we are lifting off into the future right now. And when you went from, I think I had some dsl that was basic, and then t one capabilities came into our neighborhood, and I was downloading a lot of music from napster when that happened.

Bruce Sharpe:

Yeah, same here. I remember doing remote support on dial up connections. I mean, I'm sitting here in my home office today, and I've got a fiber connection to my house that I'm running 1.3 gigabits. Him too. So, I mean, the world has definitely changed immensely.

Demetrios:

Yeah. Yeah. So, all right, cool. You became the patent expert. Basically, people would come to you if they had ideas, you would patent it or help them patent their ideas. How do you even go about patenting stuff?

Bruce Sharpe:

It's actually pretty challenging. You have to start obviously with an idea, but with the idea, then you have to really blow it out a bit so that there's a good level of understanding and a good use of technologies, et cetera, within it. But it's really all about taking that idea, fostering it, growing it, and then you've got to articulate that in a series of documentation. So whether it's a written document or a document, mostly with a lot of images, you know, flowcharts, diagrams, things of that nature, you can then take that to a patent attorney and look at viability. So from there, you know, there's a whole process. You can do a provisional patent, which is a kind of a temporary patent that you can file that's good for a year. So if you aren't sure that you really want to go to a utility, which costs you a whole lot more to actually file with the USPTO, you can put in a provisional. It'll give you some time to think about it.

Bruce Sharpe:

If you decide to do that, you can convert it to utility. Otherwise you just let it expire. But what it does is it sets up punk and time in the sand that says ten months later. If I decide to convert it, I get to use the date that I originally set back in that provisional as the date for that utility patent. So anybody that tried to come after that date basically has to cite me.

Demetrios:

No dice.

Bruce Sharpe:

And any changes they do to try and overcome my patent or their infringement, you know, they're infringing on my patent, so.

Demetrios:

Fascinating. Eventually you made your way to the AI space, and I feel like you probably didn't it didn't take you that long to pick it up. What was the antithesis that brought up my pilot that you're working on now and maybe explain a little bit of what it is?

Bruce Sharpe:

So I, you know, over the years, I did a lot with virtual agents. So when Tim Houlne, our CEO, actually reached out and recruited me, his mantra was really, we've got to have the best digital agent, right, because he knew the world was going to transition from only live people to having a digital arm, right, to support whatever the support may be in any business. So in a lot of business, it's really kind of a combination, a mashup of the two. But anyway, in the digital side, he really wanted the, the smartest, the best agent, the best sounding agent, right, that could actually converse either through a text channel or through a live voice because you knew voice wasn't going to go away, right. Even today, there's a lot of people talking about voice being the predominant channel still, that a lot of people don't trust a machine. They trust a human. Well, you start getting more trust when you start getting a better sounding voice, one that has not only a good sound sounds real, but it also has some empathy to it. We did a lot of work in evaluating technologies.

Bruce Sharpe:

Deepgram, who you mentioned earlier, is one of our partners. They've got a really good sounding technology, really does well for us. It's a part of this digital agent strategy that we have. But that's really been my focus is the digital agent isnt just enough to build a digital agent, because its all based on whats called an intent model. And intents are, what is your intention when you're reaching out to me to get support from my company? Your intent might be to open an account, to close an account, to get support on something fraudulent, whatever the case may be, then you're asking this digital agent to understand that through what youre your utterance, what you're speaking. So it has to understand that really well, the speech to text. But then you are building a lot of these flows behind the scenes, rigid flows on how to interact. Well, that's where some of the AI has really helped out a lot, because now what you can do is just, even for simple things like general questions about a business, I can actually, now, through some technology that we built, take content from that business, and I can ingest it.

Bruce Sharpe:

And then I can use a language model. I can query it, use a language model to actually compile a response and then pass that back to the customer so that you have an authoritative answer based on the content of the company.

Demetrios:

Right.

Bruce Sharpe:

So it's not as good as a light person, but it's close enough to be able to answer all those simple, easy questions and in a lot of ways, some complex questions. Right. I with authority. And then what we do is we use the live agent as a backstop in the event that you get somebody that just still is so standoffish about trying to deal with any kind of computer that they can overflow to a live agent. Or in those situations where you've set some rules to say it's gotten too complex, you can actually pass that over to the live agent as well. My pilot is really a culmination of a lot of these capabilities. Having the digital agent, the live agent, having AI to support any of the inquiries that the customer has, having AI in the conversation, divulge content and direction to the live agent when they're having a live conversation. There's a lot of different avenues and a lot of different ways that we're actually leveraging AIh.

Demetrios:

So one thing that I know can be very difficult in this specific use case is when you have a low margin of error, like you're in a regulated space. If it is an insurance provider that you're giving support to. Right. That insurance provider support can't tell you the wrong thing when you reach out to them because there's potentially people's lives at risk, or at the least, there's a lot of regulation around it. How are you making sure that things don't go off the rails?

Bruce Sharpe:

Well, we're definitely doing a lot of testing in the configuration. There's many parts to some of the AI. There's not only the content itself, the underlying base model that you use to leverage, but there's also the prompt that you actually use to drive the question right from the individual. The bigger problem has really been for us, people not wanting to have their content trained or training another model. Right. So our technology, we're actually creating separate tenants, and we're isolating that, using our cloud partners to be able to make sure that nobody from the outside world except the customer, through one of these mechanisms, a digital agent or through the live agent can actually get to that information. Right. But another thing is a lot of testing.

Bruce Sharpe:

So we're testing the content, we're testing the prompts, things of that nature. But we're also putting in logic that says for these types of things, the digital agent will never respond to. We will not give answers. We'll actually hand that off to a live agent because there are a number of things, whether it be healthcare or finance, etcetera. And it's not so much regulatory. It's really around the things that you're talking about when you're putting somebody's life in jeopardy or their well being in jeopardy, right. On one hand, it could be somebody with a very bad illness. On another hand, it could be somebody who needs something from their financial institution, otherwise they're going to lose their home.

Bruce Sharpe:

Right. Yeah. So in situations like that, those are just things that you have to pass over to a live agent. If anybody says that, you know, at least in my mind and through my experience that an AI model can handle that 100%, I would not trust them.

Demetrios:

I like where your head's at. I do not trust them either, because I think we both are on the same page. We've played around with AI enough to know that it's never 100%.

Bruce Sharpe:

Yeah, for sure. And like I said, on our Mypilot platform, we do everything that we can to make sure that the customer comes first. And it's all about the experience. Right. So that we have the right experience. So that experience means some digital, some live, then so be it. That's the way it's got to go.

Demetrios:

Well, and I like the idea of the digital part, is alleviating the burden of the live, because a lot of questions that can get answered by the virtual are, which frees up the time of those people that would otherwise be answering those questions.

Bruce Sharpe:

Absolutely. Yeah, absolutely. It makes them a lot more productive and they're actually helping people. I mean, it makes the agents actually feel a lot better that they are now helping, as opposed to just answering generic questions that could have been answered by the machine.

Demetrios:

Yeah. So have there been any large challenges as you've gone building out Mypilot?

Bruce Sharpe:

Yeah, I think the challenges have been, you know, we're not traditionally a product company, meaning that we don't build everything, we actually use partners. So in doing that, like Amazon, Google, Deepgram, a lot of them are parkers. So what well do is well go and well identify kind of a build versus buy mentality. Well go in and identify areas that we actually want to invest in, meaning build and areas that we might want to buy. And theres actually kind of a hybrid in between. Some of those we may want to build, but were going to use parts of other people's technology and then build some overlay or some other pieces to that. So I think the biggest challenge has been finding somebody who can actually do things well enough that you really want to stake your claim in their technology, you know, if you're going to partner with them. So, you know, it means that you have to do a fair bit of evaluation.

Bruce Sharpe:

You can't just go with the first guy off the street just because they're the, their name is the biggest, they've got the most money. You know, everybody's, you know, clapping and applauding them. That doesn't mean they're the best and it doesn't mean that they're the right one to partner with. Right. Because just because they're good doesn't mean that they make a good partner. So those, you know, that's been some challenges. Right. The other has been when you're developing.

Bruce Sharpe:

So we've got a very small team for development and we do some contracting. So we augment our staff with contract staff. And the challenges that you run into there is that while you can get a fantastic partner like a couple of them that we have today, you still have challenges. Challenges with staff changes, movement, things of that nature and every one of those kind of things above and beyond any problems specific with what youre trying to overcome with the technology. Those types of things can put a detriment in timelines and cost and you might miss targets and goals and things of that nature. So, I mean, its a balancing act that you really have to go into and really understand that well enough that you know how to respond or react when x happens. You know, what's, you know, what are you going to do with y, right?

Demetrios:

Yeah.

Bruce Sharpe:

So, yeah.

Demetrios:

Yeah. People can be messy and that's, I guess, why there's the push to can we, can we make it more automated? Yeah, whenever it is possible. But like we were talking about, the automations help free people up to do more impacting work, more valuable work.

Bruce Sharpe:

Agree.

Demetrios:

I'm curious, when you were going through the evaluations on different partners, what were things that you were looking for? What was the criteria that you held in high esteem?

Bruce Sharpe:

It's a lot of different things. We look at how long they've been around. Is it somebody that's going to be around tomorrow or not? We look at their achievements, successes that they've had with other customers, maybe not necessarily in the same area, but at least in the context of things that they're bringing to the table, we'll look at cost. We definitely have to make sure that we're controlling our costs. There are also things that we can get into, like we've done with Amazon and others where they've got some programs that actually provide funds to help with some of the development, as opposed to you footing the entire bill, that does mean you have to use some of the services and things of that nature, but that, you know, that was your intent when you went into it to begin with, so why not? Theres the team mentality, the processes that they go through. Like if theyre a development company, do they use an agile scrum methodology or their waterfall? We need to move fast, so we dont want waterfall, we want agile. If its a partner that were going to use parts of their technology, it boils down to cost, scalability, redundancy. If we have a situation where something falls down, how do we pick it back up, how do we restore connectivity, or better yet, how do we keep it alive so that nobody ever notices that there was a fault? Anything that happened in the back end, it's just all the it people scampering, you know, to identify what happened, even though the customer never knew anything.

Bruce Sharpe:

So I mean, it's a big laundry list of things that we go through when we go through to evaluate. And even on simple things like we did an evaluation last year for a new virtual desktop platform and we went through a fairly extensive, we evaluated 16 companies, and the process that I put together, we were able to narrow that down extremely quickly, get to a final selection, and even do a live pilot with a couple of them to be able to select the vendor that we wanted to go with.

Demetrios:

Awesome. Well, I think that's it. It's been great talking to you, Bruce. I appreciate you coming on here.

Bruce Sharpe:

Cool, thanks a lot. This has been awesome.