AI Announcer (00:05.494)
may I have your attention, esteemed listeners? Allow me to introduce you to Prompt This, the podcast tailored for those illustrious business leaders who have quite simply had their fill of the incessant AI fanfare and are yearning for the genuine article. Permit me to present Greg, your distinguished expert, whose wealth of experience in scaling sales teams is truly unmatched, alongside the remarkable Clint, a veritable titan of the startup realm.
ebbed at transforming grand visions into flourishing enterprises. Together, they deftly navigate the cacophony to provide you with insightful analysis and practical playbooks, all centered on leveraging AI to ignite fresh ideas, expand your enterprise, and of course, ensure you are not left languishing in the shadows of your competitors. And now, without further ado, allow me to unveil your esteemed host.
you
Greg (01:09.602)
This is Greg.
And this is Clint.
This is a great episode today. Someone I've known for a while. And what I really enjoyed hearing was how companies culture is record absolutely everything from phone calls, meetings, customer engagement, and being able to pump all that through AI and get learnings.
really enjoyed in today's episode was the real practical advice that our guest gave on how companies should get started with AI. Some real good tidbits in there and maybe not the exact answer you might think of right away.
Let's get into the show.
Clint (01:48.428)
Let's do it.
Greg (01:55.51)
Today, I'm really excited. We have Chris Angus, the vice president of CPaaS at 8x8, where he leads strategy for cloud communications and customer experience innovation. With two decades in the industry, Chris brings a rare mix of executive insights and hands-on tech perspective. He's known for his straight talk, sharp humor, and practical approach to how AI is reshaping customer conversations. Welcome, Chris.
Thanks Greg, thanks Clint for having me. I like the intro, kind of my straight talking. That's normally a code word for something less impressive or kind of more direct. I'll take it.
No, we want nothing but the straight talk here, right? That's exactly what we're looking
Yeah, there's no digital swear jar on this recording, so we'll see how it goes.
So I've worked with Chris in the past. I worked at eight by eight and we were counterparts. And what I like about Chris was, he's the guy, every company wishes they had someone like this or are currently looking for someone. can drop into a sales situation and fix it. He can drop into a customer situation and fix it. If finance can't figure out how to build something, he'll let you know how to do it.
Greg (03:12.994)
This guy can fix almost anything in any arena. And it's been awesome having a opportunity to work with someone like him.
Hey, listen, guys, think Greg, from that, I'm not going to be using AI for my self-evaluation next quarter. think I might bring you into my self-evaluation. Yeah, you should. That was a far too generous anecdote, but I appreciate it. I wish I prepared something I like about working with you, Greg, but I just, just couldn't find anything. Oh, I got to ask what was
I should update your LinkedIn to.
Clint (03:39.864)
like to work like great. Come on, you got to give us at least one little story in there.
We shared this off camera recently, just now, and I think if it's quite true, it's like we'd get to this part of the day when we were recording, we're like, hey, how does your call go? Mine went like this. Okay. And that gave us the kind of red or green button of how one of us had to then prepare for the upcoming call with one of our superiors. It was one of those kind of like we were going to battle together. One of us would unfortunately be on the front line taking the first bullets and then the other one would be working out, okay.
What moon is the guy in? How do we circumnavigate that? Christ.
I've known Greg for a long time and there's been a few times where he'll like lean forward and you think he's going stepping in first and then you you're right behind him next thing you know you're you're in the breach first. He's got a good talent for that.
That Teflon-coated, slopey shoulder, just that all it gave it the slit.
Greg (04:30.67)
Let's get into AI here because imagine if we had AI back then
No kidding. No kidding. It's amazing, especially in the communications industry that you're in, Chris. AI is not just changing customer conversations. It's rewriting how businesses listen. And so I'm looking forward to really digging in today about how companies are using AI to bridge human and digital conversations and maybe why starting small might be the smartest move in the AI strategy. All some good topics for us to dig into.
But before we go into any further, tell us real quick, who is 8x8 and what is CPAS?
Yes. So where do I start? So it's shifted massively over the last like 15 years, but these days, I think the best way to describe us is we're a communications platform company. So we've transitioned from this marketplace where we're looking at Unified Communications or UCAS or Contact Center as a service, so CCAS into a whole platform story. So we're the underbelly of how organizations communicate both internally and with their customers. So that's a combination of
the back office unified comms, then your front of house contact center, and then the CPAS element, which is your communications API.
Clint (05:51.704)
So let's dig into that. So as you think about AI and customer conversations, what is that, that exciting shift that you're seeing right now?
I think it's the consumption-based modeling that tokenized play. think that the shift I see now is that it's so difficult these days to attack a direct ROI on putting AI from a box perspective into a service industry. It's really tough. Let's go back 20 odd years for Contact Center. Contact Center applications, it was a cost center. Difficult to put ROI into them, difficult to get investment from the IT buyers.
on the CIOs because it just absorbed money from the bottom line. It was hard to add value to it. AI is in a similar spot. Yes, it's making things easier. So think for me, the interesting part is from a GTM, from a go-to-market perspective, is the way in which consumers want to pay for these services. They're not into fixed contracts. They're not tied into numbers of licenses. Like, hey, I'm on a pure consumption model. I use this many conversations inside my AI services.
I pay X amount, but if I have a user that uses less of them, I'm not on a fixed price. And then the same, if you tie that into our CPAS, you just imagine in a retail space where you're sending out messages about updating on a new, you know, new number stock or something else, something like that. We can use that same chat stream through SMS or RCS or WhatsApp to show a whole carousel catalog. And actually they can use AI to make suggestions on what product else might fit with that. you're buying a bike. hey, if you consider the water bottle.
And you consider the helmet and you can do all of that through an SMS string without ever leaving your device and then take a payment in the same process. That's complete automated sales transaction that not a human has touched. And you may have, you may have cross-solved three or four things just from AI making it during that chat stream. That to me is really exciting. for me, that from this side of the house, it's exciting. The worrying side is that I am so easy to get.
Chris Angus (07:54.35)
If you're a company marketing to me, I mean, yeah, I need one of those. I need one of those. So I'm like, I'm also the target market on the other end.
Well, that's a great point there. Where do businesses that know they need to get started with AI, where do they start from your perspective? Where's the best place to begin?
So I'm saying for me, so this is one of things I talk about a lot is it's often, it's often the place you start third. It's your data house. So before anyone starts looking at, Hey, I want to use AI. So I've sat in lots of CIO forums where it's coming from the top down. We need to introduce AI, but I was quite sure where or what to do or the value or how to add ROI or how to associate a cost, which department it sits in. then when we get down the line of helping the customers work, okay, what's the...
What's the low hanging fruit? What is going to give you the biggest return, but it's going to have the least impact on customer sentiment to start with in case it's not quite perfect to roll out. Right. You don't want to do any brand damage by going out too big and too hard. I'm straight out the gate. Often when you get further down the line, I find that customers don't have their internal data in the right place. And we know that, you know, we're prompt analysis. it's, you know, it's, it's, rubbish in rubbish out.
The value of anything from an AI engine really comes on the skill of the prompt and the data it's pulling from. So for 8x8's internal perspective, we use AI heavily internally these days and we do a lot of work on upskilling our teams in doing so. But that's ring-fenced and guard railed to 8x8 only data. So we can put some trust behind the sources we're getting.
Chris Angus (09:35.384)
There's no hallucinations by going out off of over the board internet or making assumptions and filling in the gaps. We can really trust the data because we put strict guardrails in. started really small.
It's coming from conversations that you know because your people were on those conversations, That's the data that you can trust, right?
As you see AI coming in more more areas, you just talked about training. Are you finding that now training on AI is becoming something that's built into every session that you're having to put across multiple teams and multiple employees?
Yeah. So we, we, we do it a number of ways. So actually we, we have kind of, we have a AI council internally that make the assessment. So we have lots of AI tools and we can touch upon some of the tools we use and why I think they're really cool. but we, so we have an AI council which looks at how we utilize AI internally. They share stories across the departments, across the teams to them work. I, okay, great. We can make sure we can, we can cross function those and make sure everyone's getting access to the same tools and same knowledge because I'm sure both of you will know that you've got to.
an organisation that transitions across all of the time zones, across every country basically, really difficult to get a kind of a local feel, such a global presence. This council come together and make sure we really addressing those kind of international sort of breakpoints. But we also run an individual AI training session on weekly basis, which is mandatory. So that's from the sea level down, it's know, through fear of death. If your employees don't turn up,
Chris Angus (11:13.954)
then it's, we, you know, we give a rattle on the knuckles. It's something really important we have to go to, we have to turn up to. and with those, what I find really cool, it's not just like so difficult to do remote training and make it interactive and make sure it sticks. mean, certainly for me as a cognitive learner, have to, I have to, I have to listen, read and do before something really sinks in personally. I'm saying.
I go off somewhere else mentally and I'm, know, often I've, I've read books and I've to read the same chapter four times because I'm thinking about something else at the same time. Right. That's how my brain.
you
Chris Angus (11:51.382)
A book when I was forced to. was being generous, Craig. Thanks for calling it out. Yeah, that's it. Yeah, that's it. Yeah. was one plus seven, right? the salesman. Yeah, I read some books once. So we run custom simulators within these training sessions and then we score them and then it creates this competition within the teams. So these simulations we've got for our AI training tools.
Hells, man.
Chris Angus (12:17.998)
I can go from anything from soft skills to market intelligence to sales training and coaching and how to address our new sales methodology and how to open up A by 8 now, transitioning from two or three products from Greg Backends Outday together to like 18 products on the truck now that we can sell. It's difficult to get salespeople to be experts across that portfolio. You could argue you don't really want them to be experts across that portfolio. We want them to open up the commercial relationship and have experts come in.
Whereas our AI training tools, they coach the team to spot and identify a couple of nuggets where an opportunity might arise or where we can actually spot problems we can solve with some of the solutions available to us. yeah, it's something which is definitely going across the board on a weekly basis we run these sessions and we've really, really leaned into it. I'm actually one of the things I'm quite proud of our teams have really leaned into, they've really got on board behind it. And everyone's now starting to see the value of AI.
from its simplest form, is, know, help me write this email right through to, Hey, here's my customer transfer from this meeting. Can you spot the five things I could have done better? Can you identify five products that haven't purchased from us today? And actually you can identify the four risks to this solution over the next 12 months. Like we can go down to that.
You guys seem like you're yeah, you're you're like next level using it already. So this is that's amazing.
Let's dig into that a little bit more. I'd like to hear more about the tech stack that you're using, the AI tech stack that you're using inside the company. Can you put some words around that?
Chris Angus (13:49.208)
Yeah, of course. So we're using like an AI orchestration tool, which then like captures like Claude, Anthropic, ChatGPT, Nano Gemini from Google, all of those in one place. And it allows our users to access them through our kind of single sign-on portal. And then you can start to build your own AI assistance for those for your own individual use.
We can also share like custom GPTs between departments so that we've got like, you know, we've got them all built into one so people can access those.
So how does that work? Everybody's all the LLMs, Claude and Plexity and ChatCPT. Is there a tool that sits on top of that? What's that?
Yeah, so it's called Air AI is the one across the top. the way that works is I think, so commercially we pay for Air AI and then all of the other LLMs underneath are tokenized. So I am an extensive user. My weapon of choice is chat GPT. Mainly I think it's because it's one I was most comfortable with. So I understand the mechanics on building my own GPTs, putting my own knowledge base in there and using it for me as an individual that it works the best for me.
But you might have engineers who actually want to test their code in one of the, um, the Claude code builders. Um, or we might have somebody who finds that actually. Gemini is better for digital creation, video content and, and making image representations of custom diagrams. So I know some of our sales engineers are able to turn the cool recordings and notes, which you capture for eight bytes recording tool. They can turn them into visual representations of, um, like solution architecture bills and that kind of fun stuff. So the fact is.
Greg (15:25.548)
Yeah, what's a good example of something that just blew you away? Like, you know, once you kind of got your hands around it, started being more effective with it, and then you probably ran into something that just blew you away once you saw the output.
So I think actually, there's a couple of things. So I think when I first started getting my hands on, it was more like my language cleanup when it comes to just in when like writing emails, things like that. That's the simplest version of doing it right. So I want to start there. If that's when I first started, okay, there's value in this. I probably sat on the fence to start with because I didn't quite understand how to use it. But as you know, it's like, I'm drowning emails, trying to write professional responses or you're trying to write responses a little bit spiky. You want to go, hey,
Gemini, make this less passive aggressive, please. And then that's how it starts in, it started there and I started to see the value of this. And then I do, I do a lot of like, customer events and I do a lot of, analyst events and I do a lot of like media writing and things like that. And what I've been able to do is input a lot of my, my videos, my writing, language, and I've been able to then train my GPT to respond to things in my tone. And when I read back.
of stuff that sounds like something I would say. It sounds like the language I would use. that's almost like, it's almost like it's built a very personal version of me, which I, which I love because it just cuts down, cuts down so much of my time having to respond to these things. But most recently, one of our engineers was building a prototype for an idea for a new solution of ours. I won't go into too much detail because I'll probably get told off for sharing any more information before ahead of time. used
It's just us three, no one's listening. It's post-earning school, it's fine. I made it in school last week. It's okay. No, so, and they, when they showed me the prototype, was like, Hey, we built most of this, most of the, most of the beginning coding came from AR. And I was like, are you kidding me? Like this is the prototype of this with like in our kind of staging environment with me and a bot.
Greg (17:07.146)
Easy there, Clint.
Greg (17:27.854)
That's really neat.
Chris Angus (17:35.106)
And I was like, okay, this is worryingly good. Yeah. Yeah. So, I just, for me, that just shows like the, like, there's those are like obvious use cases for using AI and I can talk about them. The cows come home, selling it like, you know, in the portfolio that we offer to customers, not just how we use it. but that there, because even the concept of a human brain that writes code, that builds the stuff that we're using today, blows my mind.
I just can't fathom the intellect required to do that. They had to put it into a machine and then like shave off 90 % of your time in doing so from the first pass or post. the applications are phenomenal. Imagine how fast our industries are going to excel across the next 10 years with these.
Clint and I are finding a lot of that. Like we built this podcast from basically nothing and it started to learn us pretty quickly and start to cut down. It's like you said, 90 % of what we were doing at the beginning. Now it understands the podcast inside and out and preps us for these, for these calls. And it sounds like us when it's prepping it and it really gets to know you. I have a different, you know, me.
We used to talk about looking across the entire.
customer communication, bringing learnings back up and having the upper management make some decisions across that. Are you finding that AI is helping in this area?
Chris Angus (19:09.71)
massively is one of the things we've one of the areas where you invest in hugely. think the first thing first was to try and create the try and reach out, check them. again, try and change the thought processes of the, of the staff internally. So the first thing we do is like, want, we want to make sure we're trying to record all of our meetings at all of our calls. And it's not a case of somebody in the management team. We're going to watch every one of them. Frontline managers. Yeah. So we use AI to, like, this is one of the things I love about it.
The frontline managers use AI tools to listen in and give like automated feedback on the calls that we've recorded and the meetings. it's those really cool things like, there's a black spot here where you did a four minute monologue and didn't give the customer time to speak. need to, we need to, we need to work on that. Or you close the call without capturing an action point, those kinds of things. So those learnings are really good when it comes to like coaching frontline teams. And you can do it in a non-invasive way and you can do it in a way which it doesn't feel like it's.
that took too personal is if there's a pay, these are the facts and anything we can work on. And then you can create scoring metrics so they can start looking at the next calls and meetings to see how they did against those action points. So that there's that side, but we take all of that data. Like I think imagine there's thousands of people on calls, wherever you are, CSM sales engineer, solutions architect, sales person or SDR for someone or someone in channel. All of this data goes into a massive repository.
And we start to learn the, what our customers want, what upsets them the most. What's the reason the customer gives you note, actually, do you know, I don't think I'll continue with the service anymore. Or what's the point where they go, okay, I'm to take this product now. What's the time stamp? How long were they with the company and what do they buy first to when they then start making the next purchase? So all of this stuff is like that number crunch, that data crunch is almost impossible to do. But all of this stuff's coming from our internal data.
Like from our CRM, from our customer conversations, and it's allowing us to train the business on what to expect, how to model things. okay, where we need to, okay, we need to get this many products into an organization to make them stickier.
Clint (21:15.96)
So this is where you're starting inside of your own company with AI. I'm guessing this is what you recommend to your clients where they should start too? Or do you find that there's other starting points that are coming up?
I would say a starting point would be in the, I like the of the conversation analysis as a starting point. It's the easiest thing to start with without needing to make sure your data warehouse is in a safe space. So doing that level of analysis, you need to make sure that all the information that you've got on your customers in one place or place that's actionable. What you can do well to start with is analyzing conversations, looking at sentiment analysis, looking at...
real time kind of tone and emotion and agent assistance during calls. I haven't even been a fan of like live listening to intercourses for salespeople or service people. I don't think that if you're on the front line and you've got someone whispering in your ear, I don't know that that's the right way to coach in a live like techie environment. I like the idea of a visual prompt like red.
brother right a little intrusive a little a little make me nervous if I was
It's hard to execute as the person speaking when someone else is speaking and switching your, your, think what you think you should be doing and someone's talking in your ear before you're about to respond.
Chris Angus (22:32.962)
Well, imagine if all three of us spoke at the exact same time and Greg, you and I having a conversation and Clint is prompting me what to say next then Chris, don't say that, Chris, don't say that, Chris, Chris, you're losing him. Chris, say something different. Like there's no way I could do that. let's try.
Let's try it.
Live feedback.
Live barge whisper monitor demo right here.
Maybe I won't use Clint for my review next year.
Greg (22:59.835)
That makes a lot of sense. Yeah.
In part, like the sentiment analysis is super easy to do, super easy to implement. You can do it for your back office, unified communications teams, or you can do it through your front office and contact center. So you don't have to be on these super expensive kind of sometimes daunting license packages. You can do that across the board now where you can get live analysis of the call and agent prompting and scripting and dynamic coaching. That for me is the easiest way in and it's least intrusive.
and has no impact on your, no real impact to start with on your external brand. But it's a fun way to learn how to use the tools.
What's your number one piece of advice to our audience for those business professionals who are ready to get started with AI, don't know where to begin?
So I, I think I would assess one of the, the low hanging fruits. I would assess, okay, what's the highest volume of traffic in your staff that has the, which you think is easiest to tick the box. Start, start there. Okay. So it's the least complex. So that was the pain point. Then why the pain point to start with, then ask yourself, how would you improve it from a human perspective and what your desired outcome.
Clint (24:07.118)
Identified pain point.
Chris Angus (24:17.848)
be and then work out is AI the answer? Always ask yourself that because you might find it might not be. It might just be another token in the IVR on the call into the contact center. It may not be AI. So my advice when looking at AI is don't always assume AI is the answer first pass to post. Identify the pain point on the problem. Understand your ideal scenario. Work from there.
And then work with a trusted consultant or an organization that you can trust to give you advice. Who's not just trying to sell at all to you. There are tons and tons of experts out there. we're an organization that has a bunch of them in our business that will give plain and simple advice on how to provide you a solution for a problem you're trying to solve.
Great. Absolutely solid advice. Don't just throw AI at it. Understand the pain point, understand the process, then figure out the solution. And if it's AI, go with it. Nice. So let's shift gears. Let's bring it back to you on a day-to-day basis. How are you using AI personally? Is it fitting into your personal life or is it just a work tool for you?
No, so I also pay for a GPT subscription personally as well. That's because I don't want my corporate license knowing what I talk about, what I think about. If they have enough of my time, they don't need my thoughts at the same time.
I know already. Yeah, I do.
Chris Angus (25:53.326)
Yeah. Fedor was listening. I love you guys. I'll give you a real live example. So I don't want to be one of those guys, but I am running a marathon next year. I am one of those guys. It's my first one and I'm not a runner. I am not built for running at all. But I used ChatGPT and through kind of maybe half a dozen prompts, I built out a really cool training program.
And I shared it with a coach friend of mine who from my gym. and he, he was like, this is, this is pretty bang on. This is, this is really good. He's like, Hey, actually I'd like to see you taper earlier. If I was coaching you. So I was like, okay, I went back into my prompt. I was like, Hey, can you consider these prompts and can you actually consider this tapering as a separate thought process that gave me a new training program? And my coach was like, that's really good. That's no different to what I would probably given you if I was coaching you for your race.
and then I'll take it one stage further. It's like, okay, cool. Now can you give me, eating and hydration plan throughout the, marathon? So now tell me what great points I needed to take on, fluids or, electrolytes. I also asked it to ask me questions. Hey, before you do this, can you ask me five questions? I make you easier for you to give me the advice. And it was, kind of had to, Hey, you know, I think it was like, yeah, how, sweaty are you? kind of thing. How much do you perspire? How much do you weigh? These types of things. And they come back to, okay. Of your height.
size per-storation weight. Um, it was able to give me some guidance on when I could consume water, when I should consume carbohydrates at what grams per mile and all this kind of fun stuff. that's a good example of how I it. It's pretty cool. And I, I went, um, my wife's running with me and there's a bunch of us from our gym doing it and that's how I got dragged into it. Like, you know, uh, Greg would tell you from our past, you want to, you want to get me to do anything? Tell me I can't. Oh, I'm not sure if Chris can do that.
mindset.
Chris Angus (27:45.806)
Yeah. Like he's like, you know, I'm the most pigheaded competitive person you'll meet. I'm okay with that. I've accepted it. It's who I am. yeah, was one of those. Yeah, exactly. I just learned it. Right. So it was one of those conversations like, Hey, I'll just come for the fun weekend. And then, and then my mate's like, why don't you do it? It surprised everyone. I'm like, yeah, I'll do it. And then as we running through the list, as my friends are running through the list of people who doing it, every time they get to my name, like really?
The level of, the level of people is like surprised to hear that I'm doing it. I'm not sure if I should be proud I am or somewhat of spending with what I do.
Greg (28:31.502)
Okay, it's time for this week's AI Challenge. Now the AI Challenge is a takeaway assignment for our listeners to get their hands on some AI tools and do some exercises.
This week we're having a little bit of fun with it. You heard our guest talk about how he uses ChatGPT to come up with his marathon preparation regime. So what we're going to do here is have you use ChatGPT to come up with your own personal exercise plan. So have a look and have some fun with this one.
Yeah, look down into the show notes and you'll find the link that goes right to the blog. It's got all the instructions that you're going to need. Now, if you have an AI story in business to share and want to be on the show, go to www.promptthis.ai, go to the contact us, fill out the form and we'll be in touch with you.
Clint (29:29.614)
Chris, I gotta tell you, I really appreciate the conversation. It was a lot of fun. I liked learning a little bit more about Greg. It's always good to hear some Greg stories in there. Thanks for joining us today.
And if Chris, if anyone has any additional questions or wants to connect with you, where where would they find you?
So it's email addresses, chris.angus at 8x8.com or you can find my LinkedIn Chris Angus. Thanks having me guys.
thanks. Thanks for joining us. That was a great conversation.
Alright Chris, you have a good one.
Chris Angus (30:00.398)
Cheers guys.
And that's another episode of prompt this.
AI Announcer (30:11.318)
dear guests, I must extend my most heartfelt gratitude for your esteemed presence alongside our illustrious hosts, Greg today. Might I draw your attention to the splendid collection of Prompt This episodes, not to mention a plethora of thoroughly engaging articles, which may be discovered at www.promptthis.ai. And if I may be so bold, do be sure to engage with the follow button nestled just below.
We eagerly anticipate your return and it would be our utmost pleasure to welcome you back.
