Artificial Intelligence Discussion Panel featuring Bluewave, Bright, Capacity, Five9, and Rackspace

Paul Weiss, Bluewave:

My name is Paul Weiss from Bluewave Technology Group, and today we are going to have a round table discussion on artificial intelligence and more specifically demystifying artificial intelligence.

I’d like to start with a thought and that thought is gravity. What does gravity have to do with artificial intelligence? Well, let’s talk about that for just a second.

Can you calculate gravity? Do you know how? Probably not. Most people cannot, but that’s the good news because you don’t really have to know how to calculate gravity to use gravity.

Think about this. You understand its practical applications and its use cases. If you put your foot on the ground in the morning when you get out of bed, your foot will stay there until you lift it up. If you throw a ball up in the air, that ball will come down.

If you need to jump a puddle, you know exactly how far you need to jump to clear the puddle. Every single day you make thousands of movements and decisions that put gravity to its practical use.

Well, welcome to artificial intelligence. You do not need to understand how it is built or the mathematics behind it or even how it learns. You only need to know how to use it, in what ways, and to what benefits for your work and for your organizations.

Today we are going to be introduced to four very intelligent and wonderful people that represent innovative providers delivering AI solutions in the marketplace. And we are going to discuss how they offer simple and effective solutions that help organizations leverage AI to make human work easier and to drive positive impacts for both our clients and our businesses.

I will hand over our first introduction to Michael at Five9.

 

Michael Roche, Five9:

Hey, all. Thanks Paul. Appreciate that. Wonderful and intelligent people. So again, Michael Roach here from Five9, RVP of Solution Consulting. I head up the AI portfolio of our product set. I manage a group of solution consultants, so kind of on the technical sales side of the house. Tim?

Tim Yeadon, Capacity:

Yeah. Hi everyone. Tim Yeadon with Capacity. Capacity uses AI to automate common support tasks, whether those are internal tasks to support your employees and team members or external support customers. So looking forward to today’s discussion. Pass it off to Luc.

Luc Dullaire, Bright:

Hey, thanks Tim. I’m Luc Dullaire, I’m head of growth at Bright. Bright is an immersive learning and simulation company. We use AI powered simulations along with a whole slew of other capabilities to deliver practice at scale for companies worldwide. So it’ll be really fun, Paul, to dig into this stuff with you. Jeff, take us home.

Jeff DeVerter, Rackspace:

All right, my name is Jeff DeVerter. I’m the chief technology evangelist for us here at Rackspace Technology. And at Rackspace we’ve been using technology for the past 25 years to help transform organizations to do what they do better. Lately, of course, in the past year we developed a spin up inside of our organization, a division if you will, that’s focused solely on helping companies adopt AI to really make their (HI) human intelligence better. So anxious for the call today.

Paul Weiss, Bluewave:

Excellent. Well thank you all of you.

Our format is going to be open conversation. I have several questions prepared for the group and for certain individuals. We will get started on that and I can’t wait to hear the wonderful answers that we get back from our panelists.

I’ll open this one up to the group and I’ll ask Jeff to start us off with his response. Large language models, everybody’s heard of OpenAI and Chat GPT, they’re quite the hot topic, making a real boom in both business and in everyday life for just general consumers.

Are these AI models really going to fundamentally change customer service landscape and how businesses work and make decisions or are we overestimating their impact?

Jeff DeVerter, Rackspace:

No, I think they absolutely are, and in fact, I would substantiate that by saying they’re doing it today. Absolutely, they’re doing it today. And for those of you who don’t spend your time, like a lot of us here do, researching all of this stuff and reading what happens every single day. A large language model is really that brains behind the scenes that’s been trained, taught if you will, on a set of data to be able to then make decisions on that data and to be able to then either converse or translate or do a thing based on whatever that question is.

And there are different language models that do different things. There are small language models now too that are more portable. It can be used in a variety of different ways.

And what’s happening now is these language models are taking more and more sets of data in very unique ways to be able to provide better customer service, to be able to provide information in a just-in-time basis and to be able to, well in some cases, make some stuff up if it hasn’t been trained well.

And so that’s the other big thing that hopefully we’re going to dig into here today is the fact that an AI solution is not a set and forget. There isn’t one that exists today because knowledge changes and those models grow and mature over time. Once you’ve done the AI project, you’ve got to get in and continue to spend time with it to ensure that it is still relevant and is still providing the information that will continue to transform businesses.

Paul Weiss, Bluewave:

So continuous learning and fine tuning. I mean, who here stopped learning after third grade other than me? Of course nobody here on this panel. But Jeff, you talked about a couple concepts and so I’ll share this as I throw the question to Tim.

AI at a high level does these three things. It automates previously manual repetitive tasks. It thinks or scans or analyzes faster than we can to recognize trends in big data and offer insights and it generates content in real time, to your point, Jeff earlier. And we’re going to talk about how many use cases fit into any one or multiple of those three things. Tim, same question to you regarding AI and is it the big boom and disruptor that we all think it’s going to be or is?

Tim Yeadon, Capacity:

Yeah, I’ll follow up to Jeff and say I agree. It already is happening. I think what’s important for companies like the ones represented on this call is we have to find and deliver useful applications of the technology. We can’t get stuck in cool YouTube demos or whatever we see online. Our job is to work with our customers to solve real problems for them. So Paul, I like the definitions you just shared there. I’ll give a simple example of a very common use case that we’re being asked to solve: email automation. We still have to respond to a lot of emails as we’re sitting on this call.

Paul Weiss, Bluewave:

Are you saying that I don’t have to respond to all of them by myself anymore?

Tim Yeadon, Capacity:

I’m saying that that is a great use case for large language models and generative AI is not only can it assess the context of the email but actually draft that response based on your past behavior and how you’ve responded to similar requests in emails in the past. That’s a great useful cost-effective application of large language models in generative AI.

Paul Weiss, Bluewave:

Fantastic. Luc, how do you feel about that?

Luc Dullaire, Bright:

Yeah, I mean it’s hard to disagree with what you guys are talking about right now. I’ll even take it a step further because yes, automating and decisioning and generating all these things. But one of the things that we’re using AI to do is on the learning side. Jeff, you said HI. I hadn’t heard that before. I like that a lot.

But the human intelligence capability needs to continue to grow because the things that are automating are the easier things, and those more difficult things fall, such as some people still want to talk to humans and other things.

So when you think about training employees to be better equipped with these things. We can use AI and generative AI and natural language processing and other things to help scale the performance coaching function in the practice that people get so that they become that much better, that much faster.

Because no longer do you need to maybe sit side by side with somebody and train them. You can actually give them an asynchronous digital experience that uses AI to coach and provide feedback and upscale people through practice, which is how we’re sort of leveraging that capability. So in that sense, it’s a big deal on that side of the equation to continue to uplift the whole CX space in general.

Jeff DeVerter, Rackspace:

And Paul, one of the things, maybe even to add to this. One of the things I think that so many people miss is that AI and the application of it, it doesn’t have to be landing a rocket on Mars (and they use AI for that). But it can be things like making your people smarter. It can be answering emails. And when you could give 10 minutes, the reality is almost 50% of a knowledge worker’s day could be given back to them with appropriate use of AI.

What could you as a business leader or a team lead do if you got 50% more productivity out of folks and happier people, they’re not doing the things they didn’t like to do before. So the challenge really goes back to don’t think you got to land a robot on Mars to use AI. There are plenty of productive use cases today.

Luc Dullaire, Bright:

Yeah, and actually Jeff, I really liked that. Paul, you talked earlier about your gravity thing. You don’t have to know how it works, but if you know how to maybe define what good looks like in your own words and have companies like ours do all the work in the background to then translate it to something that gets useful. A practical application. Tim, you talked about solving problems. If you can give somebody the tools to do it themselves, to put some information in and spit out something that they have no clue how it got there, but they know it’s the right thing to see and do and it helps them scale their business or scale their training, that becomes the key. Making it easier to use AI to solve the problems that we’re talking about.

Paul Weiss, Bluewave:

And a lot of our, and I think Tim touched on this before, our mandate and working with our clients is to also recognize and identify those problems that do exist that do occur in business every day. Sometimes down in the work every day without the, let’s say the perspective of big picture and scale, you might not realize that these problems are at the level of friction and occurrence and frequency that they are within an organization. And so before we can magically point the power of AI at problems, we have to first recognize those problems and game out how to solve them and how AI may play a role in doing that critical in the work that we all do with our clients.

Michael, I’m going to give you the next question to get us started. And I think Jeff mentioned HI, human intelligence, and I have in the past referred to that as AI, actual intelligence instead of artificial intelligence, but it’s all the same. It’s this framework of better together. This augmentation of accelerating or deflecting human work and figuring out where and how to harmonize the work and the talents and the skill sets of humans and their actual intelligence and artificial intelligence.

What do you see, and we’ll hear a little bit from everybody here whose platforms and services kind of pointed at some similar but also different use cases and where work gets done. But what do you see as some of the most common and effective ways that AI platforms are enabling this framework of better together?

 

Michael Roche, Five9:

Yeah, I’m happy to comment on that. Before I do, I just wanted to go back to the last question just for one quick comment.

I think the other thing that’s really interesting about where we are today in terms of this evolutionary technology is that it’s democratized. Our customers’ customers can touch it and feel it now. Practitioners don’t need to be developers. They can be people like you and me who are just touching and feeling and testing and working with AI in ways that have never been possible before large language models. To me that’s really exciting because it’s creating a whole new breed of people that are involved in the AI process. But I guess we could go on forever there.

In terms of what I’m seeing. So for Five9, we’re a CCaaS platform, so we are helping contact centers and managing that volume of the front door.

There are a ton of different perspectives that you can look at this problem in terms of solving. If you’re trying to solve a customer problem, which are things like deflection or high hold times or things like that, there’s one set of potential solutions. If you’re trying to solve agent productivity or optimization, there’s a whole other slew. We’ll do both of those real quick.

On the customer side, look, the front door is the contact center and typically that’s intelligent virtual assistants or digital virtual assistants, both of which have immense power. augmented now with larger language models. to take things off the table. More things than we ever have. It’s an easy place to start. It’s really effective. And the tool set is broad, but calls do escalate to agents. And when those calls are sitting with agents, you can use things like agent assist and large language models, summarize calls and contacts in milliseconds, which is a real practical example of a solution. Typically after-call work takes 45 seconds a minute. That can all just go away by leveraging a system like Agent Assist and summarizing those calls using large language models. So again, we all have a new tool in our tool bag and it is fun to explore new areas to optimize.

Paul Weiss, Bluewave:

So Luc, for that question, I like to think about, since we were talking specifically about the agent in the contact center. I’ll stay there for a quick second.

I like to talk about that from the perspective of before, during, and after. Before are all the things that we can do to prepare, align, skill match, recruit, hire, and train an agent to be best equipped to perform for the during event, which is the actual interaction with the client.

And then after all the things we might do to analyze coach and perform performance manage and do so in a way that is meant to improve each individual agent in the way that individual agent needs to be improved, not all obviously a monolith. So as we think about from the framework of better together and that lifecycle of before, during, and after, where are you seeing great impact with AI and particularly with what Bright is doing?

Luc Dullaire, Bright:

Yeah, I guess from a couple of lenses, I think that’s a great question. Obviously before, we’re a training platform at our core. So all of the onboarding, even the new hire, pre-hire assessment process, but getting them up to speed to be the most competent and confident in the least amount of time. And Jeff mentioned something that is really important. When there are people who are invested in, who get a chance to go practice their craft to the point where they feel good about it, they don’t quit as much. They don’t churn as much. They are happier. We see that day in and day out. We have people tell us, no one’s ever invested in me. And that’s what it is. It’s an investment and this is what technology helps us enable at scale. So before, you get somebody to be very good a lot faster than it used to, and the business wins a lot from that perspective.

Paul Weiss, Bluewave:

I want to pounce on something you just said there. It was impactful. Counteracting the fear of, oh, you mean you’re trying to replace me, to this, nobody has ever invested in me. Think about the power of AI in changing that paradigm and the way folks realize that the work they do is high value and being recognized as such. We’re trying to unencumber you with some of the menial repetitive task work, but boy, the work you do is valuable, and we are investing in you.

Luc Dullaire, Bright:

That’s right. Yeah. And I mean if you think about we’re staying in the contact center, I guess as a great use case. That’s a hard job. The training is traditional learning and development is classrooms a bunch of fairly traditional experiences, but that’s not how adult learners actually learn. You learn how to play guitar by picking up the guitar and practicing it. Yes, you have to watch the video on how to put your fingers where they’re supposed to go, but you have to do it and practice it. And the ability to do that really speeds up the process and gives a lot of those things. But then even afterwards you go out and you do your job and there’s AI that’s doing things like automated quality and call assessments and speech analytics that are understanding what’s happening out there and identifying Paul’s opportunity areas versus Tim’s opportunity areas.

Maybe they’re different. Well, if you take that type of information, you can automate the upskilling journey of those experiences using AI and some other types of automation where now you’re going to automate that process. But when you think about the better together, a lot of that content is still sort of designed by the humans who have both the art and the science behind what they’re trying to teach. And AI is able to wield that content more effectively if you think about it that way. So there’s a big better together theme from that perspective.

 

Michael Roche, Five9:

I love the idea that we are focused on agents in a way that we never have been. I mean, it just wasn’t a talk track not that long ago. But imagine a world, for example, where you’re using large language models to craft an after-call summary email. So the agent doesn’t have to do that. They don’t have to draft that email anymore. Just as a simple example, that makes their life a lot better. So the possibilities are endless. And I love the focus on agents.

Paul Weiss, Bluewave:

Yeah, it’s a great use case. But there is menial, repetitive, wasted work in everybody’s work every day across all verticals and all job functions. And so Tim, I want to give you a little spin off of what Luc was just talking about.

We’re saying I’m being invested in, I’m being valued and not being replaced. As a support automation platform with Capacity, you can make the same argument that we are valuing the work you’re doing because we’re taking the things that don’t really require your skillsets and your personal capabilities and energy and power, and we’re giving it to AI so you can be you. Tell us a little bit about how you’re doing that in some real obvious places where you are lifting that work away from the folks that frankly may not want it in the first place.

Tim Yeadon, Capacity:

Yeah, Paul, we often talk about our stats around containment or deflection, meaning when we deploy our solutions and they’re utilized on a for every 10 inquiries requests, our system can handle typically nine. Well, why is that? Jeff talked a little bit about it before, right? 50% of a knowledge worker’s day is kind of frankly wasted on these repetitive tasks. And in a lot of roles it’s 80 or 90%. Let’s just say it is what it is. And so I think what’s happening is it’s preventing folks from those roles, from upskilling and moving up in the career path because the perception is the workload is so heavy that we can’t move you into a different role because someone’s got to respond to all those inbound inquiries and responses. And so I think there’s a couple of things here.

One, I do think we can help people upskill and focus on more complicated complex cases and scenarios. And then two, I think everyone on this call will agree, there’s a whole new set of jobs that is required and being created through this technology. And so we shouldn’t ignore the opportunities we create within a career path, but the opportunities we’re creating for a whole new career path based on the knowledge you have from your current career path. And so I think that’s a big part of, I like what you’re saying, demystifying this. We shouldn’t look at this as a job stealer, a job replacer. I think we should look at it as an enhancement to the employee experience just like it is for the customer experience. And we need to embrace that more as an organization. So for us, it’s always AI plus a human. It’s always those two sides of the equation to deliver useful cost-effective solutions that are going to make an impact for a business.

Paul Weiss, Bluewave:

So Jeff, for an organization that helps companies develop their AI strategies and deploy their own models, you oftentimes may get engaged in a conversation that an organization has brought you in with this and points you at a specific set of challenges. Or maybe they come to you, shrug their shoulders and ask you, where do I start? And they may have also questions about governance and responsibility with leveraging AI and transparency. How do you approach that conversation? How do you start that conversation with an organization that has little to no AI experience but certainly recognizes its potential power?

Jeff DeVerter, Rackspace:

Well, we have those conversations all the time. I was literally just on a call with one of the largest companies we’ve never heard of, and they’re in the construction world. And they’re looking at AI to help figure out. I mean the challenges they want to solve for is this:

They have literally somebody on engineering wearing a hard hat out in the field, raising a cement wall, and they want to get the right anchor point locations. They’re calling in engineers now and they want to actually use AI to pull from the authoritative elements of their documentation of where to put that sort of information. But it’s a good example because of how we have these conversations. Not Rackspace, you know us as the server company, we’ve been around for 25 years, and oftentimes we’ll walk into a room to have the conversation and people will scratch their heads and go, why is the server company here?

Well, we’re here because I mentioned it earlier, we are focused on and have always been focused on helping companies utilize technology to transform their business. And no technology can do that more than AI. So we spun up this organization, we call it FAIR, the Foundry for AI at Rackspace. And to get out of your question specifically that word foundry in the naming of our division is important. Think about a foundry back in the beginning of the industrial revolution. It was a place where artisans came together to take raw materials and turn them into something that’s useful that can be used at scale. And that’s exactly what we’re doing with AI. We have these conversations with companies, it’s asking questions like we ended up in the call center earlier, and it’s a great place to get stuck because we’re talking about how do we help these folks who are artists at helping companies deliver service to their customers?

 

There’s no better job function in the organization, but why don’t they like their job? They’ve got to do the menial stuff. And when AI can take that off their shoulders, and we have a conversation that talk about what are the things that are repetitive that you have to do or your organization has to do or your customers have to do that we could take and let a machine do? And then the light bulbs start to go off and it takes just a little bit of prodding. There’s a little bit of inertia because people have this, I have to land a rocket on Mars inertia of what AI means. And it isn’t that, and it’s not the terminator. It’s writing up follow-up emails or notes inside of a record, or gauging sentiment, or doing all the things that are inside of a call center that no one likes to do but has to be done that are tangential to providing services.

Paul Weiss, Bluewave:

So again, it’s that we talked about demystifying. It’s not about understanding AI and all of its permutations. It’s what are my problems, what are my challenges? It starts right there. That really is the conversation.

Jeff DeVerter, Rackspace:

In fact, the way that we even approach a company adopting AI is we say, look, I want you to consider this solution, this technology that everybody thinks, oh, it’s another server we’re going to throw in a data center. Now I want you to think about AI. It’s an employee that you’re bringing into the organization. How do you introduce it around? How do you govern it? How do you manage it? And when you start to think about AI through that lens, well then it starts to make a lot more sense about, oh, it’s a tool that we can use to make our lives better.

Paul Weiss, Bluewave:

I do like the idea of the visualization of the bot on your shoulder, like Kazoo from the Flintstones, and that bot is there to help you get your work done. And it again, demystifies it is it’s just another resource or utility or an artificial person that you can point at work.

Michael, where are you finding in your conversations with organizations across many verticals, and you can get specific or be broad here, where are you finding organizations gravitating to as a place to start with their AI strategies?

 

Michael Roche, Five9:

Yeah, I mean, I’m going to go right to where Jeff just was honestly. I mean, the problem is we get a lot of customers that come to us, and this idea of demystifying is really important because it’s very confusing. The landscape is challenging out there right now. There’s a lot of noise and so forth. So finding those tactical, practical places to start is typically where we find success. Just getting folks to get used to the idea. Let’s take this one step at a time, crawl, walk, run kind of approach. Let’s find the opportunities that are easy to invest in that kind of low hanging fruit, and we’ll just go from there.

But it’s a journey. It’s a longer roadmap from a product standpoint. I think I mentioned the DVAs and IVAs and then agent assist. Those are two areas that are gaining a ton of traction right now, but that presupposes that companies know where to look, which brings us probably to a third area of opportunity, which is data mining, right?

Large language models in genAI right now can do so much quickly and just ingest an enormous amount of data and reveal for our customers where they should start, where the best opportunities are, the lowest hanging fruit, the things that are impacting their business in a negative way, the bottlenecks that can all be done in ways today that just literally weren’t possible prior to November of 2022.

So for me, that’s another area. We get a ton of traction providing visibility from a reporting and data analytics standpoint to those folks that manage contact centers. And then that builds a roadmap, if you will, in terms of where to go next.

Paul Weiss, Bluewave:

Well, it enables decisioning on actual data, instead of intuition.

I want to stay there as I transition to Tim, because you hit on something that was really important. So you talked about AI analyzing data faster than we can, with bigger samples. And we know that most contact center organizations are operating on a sliver of observed data with a lot of blind spots. And certainly first, the AI can help us clear the blind spots by analyzing the data in the areas we’re not aggregating or collecting it at all. But most organizations, I think what can stipulate, are operating with a set of tools today that allow them to gather, organize, collate, and visualize data, but that doesn’t provide the insight and it doesn’t suggest the next best action and generative AI can.

And so the question of why are these things happening? I can see them in my dashboards and my reports, but where’s the causality and how do I solve for that? At present, most organizations are not operating with a utility that helps them get to that level of next best action, and AI certainly can do that.

So Tim, as a generative AI platform, help us with that. Talk a little bit about, we’re going to transition from where are organizations starting with AI to what’s the purpose in the first place? It’s to have success with AI. Where are you seeing organizations having success implementing AI in some of the ways we’ve talked about today?

Tim Yeadon, Capacity:

Yeah, so Paul, you brought a very important point there, and that is organizations are sitting on a lot of data, and I would say last decade, decade plus, my colleagues here may have a different timeline, I think companies have been focused on capturing and organizing that data for a long time now.

But the insights, as you mentioned, are really tough to draw out of it. And then once you get the insights, what action should I take? That’s what is so exciting about generative AI, right? It can help you complete the next steps in that path as you’re talking about.

So couple of areas. One in financial services, one in healthcare, but still pretty much related together what we call recommendation engines. So let’s say I am in a healthcare organization, and someone gives me a number of symptoms and I want to confidently recommend exactly what steps they should take based on their symptoms, preexisting conditions, et cetera. That’s something where we have a lot of data sitting on the backend. But to take someone who might be either new to the organization or working with someone who might be the patient directly and actually confidently providing next steps, not diagnosis, not a diagnosis, but a next steps on where to go, what to do, how urgently you should prepare or pursue those next steps. That’s something we’re working on.

Same thing in financial services. There’s a lot of financial services products out there. Which one’s right for me? Which one do I recommend to my client? I have all this data that tells me what those products are. When I think the inputs from the person to client I’m serving, it can actually help me recommend the path to pursue. And the key in these use cases is you have to have the answer key. We have to use these tools to understand what is the best outcome based on those inputs.

We talked about people and AI. If I’m in my customer’s shoes or any organization who might be watching this, I wouldn’t necessarily allow or trust the AI to come up with the best possible answer key. I would really want to supervise that process and make sure the AI is arriving at the best possible outcome conclusion. So that’s just another one.

We go back to the supervised people and AI conversation, it ties in there. But the idea to recommend next steps and any process that we face as consumers or employees or whatever it might be, I think is a really exciting use case in application for the technologies we’re talking about.

Paul Weiss, Bluewave:

And humans still have to and are always going to be responsible for bringing the emotional intelligence and the creativity to the conversation as well. And so it is, again, back to our original theme, better together is for sure the better approach.

Luc, we’ve talked a lot about, and I want you to pick up on something that both Tim and Michael actually mentioned, the completing the loop component. So we talk about BI tools as organization and visualization tools that lack the next step of insight and action. Well, let’s talk about those analytics tools in the context of maybe a contact center where I’m analyzing the work that people are doing, I’m recognizing a need to perhaps coach and bring somebody along in a particular situation. But then what? But then how do I execute that? How do I complete the loop so that I go from recognition to actual outcome or result?

Luc Dullaire, Bright:

Yeah. Yeah. I mean, today a lot of companies don’t even have the tools to do that part, right? The analytics tools that sort of overlay and pull in all the data and do all the analysis. And Tim, to your point, all of the definitions around what are the thresholds you need to be hitting from a production perspective and a performance perspective, et cetera. And then therefore, who’s falling short of that and who might need that to be coached from a next best action perspective? A lot of companies don’t even have that.

The ones that do that happens and they go and deliver it to the leadership team and say, here you go guys. Here’s your list of people to go coach. Go find some time wherever that exists, which it doesn’t. Sit with them and coach them, which many managers are not necessarily coaches or trainers, right? We’ve seen that dynamic, especially in the context of more space.

So imagine if you can take that exact data, which I think I touched on this earlier. We can ingest it for example, at Bright and say, okay, well what are the skills that you’re falling short on that you need to be coached on? There is a bank of content and simulations and practice that is available both as a business and as we identified those are all skills tagged. What skills do you need. Automatically put that together and deliver it so that Paul or Tim or Jeff or Luc or Michael can have all of the practice around their specific opportunity areas. But now I’ve completely automated the process of delivering that, identifying it, curating it, delivering it, and then all somebody’s doing at that point is confirming that those people have practiced because their performance is being measured and monitored also by AI.

So we’re rating and coaching and scoring them based off of all of the QA rubrics, Tim, that you said, humans are still the design and they still have what good looks like. But AI can automate and manipulate and manage the process and the guts of that so that when you’re going through and seeing, instead of, Hey, I’m sitting next to you, Paul, I’m going to teach you what you need to do better. I’m going to now sit next to you Paul, and say, Hey, I see that you’ve been practicing and that you’ve reached the proficiency levels here. That’s awesome. Do you have any questions? I’m going to coach you. I’m going to manage you instead of I’m going to train you. It’s a very different dynamic, and I didn’t have to spend all that time doing that with my employee. That’s one of many use cases. That’s a really cool way to close the loop from a continuous improvement perspective.

Paul Weiss, Bluewave:

So coaching a defensive end to cover a wide receiver downfield is not maybe the best use of time coaching that defensive end to rush the passer perhaps is. And so knowing who you’re training, why they’re training them, and what your charter is for improving that individual’s performance is tantamount to success.

Luc Dullaire, Bright:

Yeah, absolutely. You got it. That was a, I mean, mic drop right there.

Paul Weiss, Bluewave:

Ha. I’d like to offer this up to the group, and we’ll start with Michael. We’ve talked about where we get started. We’ve talked about what success might look like, augmentation better together, but there’s a lot of paralysis by analysis, a lot of uncertainty. Organizations that either are suspect of the promised outcomes or simply don’t know where to start. What are you seeing? And this, we’re going to go around the horn. This is a really important question. We’ll go around the horn from Mike to Jeff, to Luc, to Tim. What are the biggest factors that are stopping companies from utilizing AI? Is it that lack of awareness of benefit, uncertainty on how, where, and why to use it, governance and compliance concerns, some mix of all of those. What has been your experience to date?

 

Michael Roche, Five9:

Yeah, I think I mentioned it briefly, but I’m going to reiterate maybe a little bit stronger. And again, we’re in front of customers all day every day, and it is just the current state of confusion out there. If you’re a customer, you’re looking at a contact center vendor, there are so many technologies and so many kind of AI vendors. If you go to Contact Center Week or you go to Enterprise Connect, every single vendor is leading with AI, not most, every single one. So navigating through that can be a challenge.

So I think the best way to approach that and where we kind of instruct is let’s start with your problems. And that has never changed. That has always been the case. Let’s just start with the problems that you’re seeing and then we’ll back into the tools that we have at our disposal, many of which we didn’t have a year ago or two years ago to solve those problems. So it is just guiding people not to get wrapped around the axle with all this noise, and let’s just start slow and work on how to solve the problems that you’re seeing every day.

Paul Weiss, Bluewave:

So I was just notified. I don’t know if it was by a person or a bot, but we are coming up on a hard stop in a few minutes. And so I would like to get this question out for the team as our final. And the question is, and I’ll start with Luc, what is the biggest takeaway that you want our viewers here today to leave with as they think about and prepare for how they can utilize AI to benefit their organizations?

Luc Dullaire, Bright:

One, wow, that’s a hard one. Yeah, I think it’s been the central theme of all of this, which is really, you don’t necessarily need to know how to build it and how it works when you’ve got companies that sort of do that for you. What are the challenges you’re trying to solve? And then leverage the tools that are right to solve it. AI may or may not be the right tool to solve X challenge. Identify your sets of challenge. Mike, you just talked about what are you trying to do? And AI may or may not be able to automate. It may not be generated. You used some cool words, Paul, a little bit ago around what that looks like. So I think the one big thing, just on your last question I wanted to point out just real quick is governance is a crapshoot right now. So many AI councils are not quite sure what’s happening with data because of open source and closed source and all that stuff. So everything is compliance related. What are you doing with my data? How is it being used? So I think the other thing for us as an industry to do is better define that process, I think too.

Paul Weiss, Bluewave:

Yeah, I mean from a technology vendor perspective, those disclosures, is it a closed LLM? Is it my model? Is it a public model? Is your data training it, is your patient or customer data interacting with it? These are all very important and necessary questions.

And in fact, when we talk with organizations about evaluating solution sets, again, it starts with problem solving. What are we trying to solve for? And then the alignment of solutions, because AI can be pointed as we’ve discussed in many, many directions. But one of the very early considerations before we get too far down the line is to have a governance and InfoSec conversation first and foremost. Because we can solve for all the challenges and point all the tools at all the right problems and if an organization ultimately isn’t comfortable with the security governance and compliance aspects, it’s going to get shot dead at that point. So we might as well get that out of the way first in terms of how we begin our evaluation processes.

So takeaways, biggest takeaways, I’ll give this to Jeff now. Jeff, your question again, biggest takeaway?

Jeff DeVerter, Rackspace:

Yeah, I’ll keep it simple. So what are the areas that Rackspace focuses on is ensuring that companies are deploying responsible AI solutions. Whether those are AI solutions that are baked into an off the shelf offering. You use Salesforce, they’ve got AI in there, your UCaaS environment, there’s obviously lots of AI in there. Utilize those fit for purpose environments when you can, but make sure it’s responsible AI.

And we define responsible AI in three areas. First is symbiotic. AI is here to help humans, not the other way around. Two secure. You’re bringing all your data together, it better be secure. And then three, sustainable and think about sustainability through two areas – green AI can be very costly on carbon tax and is that it’s here to ensure that as an offering and a solution that is sustainable over the long term and raises people up. Remember back to the symbiotic thing as opposed to excluding certain folks to ensure that everyone has access to the technology.

Paul Weiss, Bluewave:

Thank you, Jeff. Michael?

 

Michael Roche, Five9:

Yeah. I’m going to take a slightly different approach. I think one thing is more of a caution as we go down the AI kind of roadway here. Be careful not to wed yourself to a singular technology. Like in Five9, we have a platform that’s open. What we mean by that is as new technologies emerge, we want to be able to take advantages. We will bring those into our fold. If you lock yourself into one technology, you won’t be able to take advantage of things that emerge. And in this landscape, things are happening really fast. Failure to do that is going to be a problem. We really want to futureproof our customers by making sure that we’re constantly evaluating new tech and bringing it into the fold wherever possible. So just be careful as you’re kind of wetting yourself to technology, not to lock yourself in too much.

Paul Weiss, Bluewave:

Great point. Not to confuse open with, not secure, right? Open in terms of its application landscape and its ability to integrate and work with tools any evolve as your work and your needs evolve as well. Thank you, Michael. Tim?

Tim Yeadon, Capacity:

Yeah. I’ll add another up vote for governance and security. Really, really important. And a lot of organizations are trying to navigate that, and it’s worth the investment of time to really figure out what you need in your industry and for your company. But I think in addition to that is. New tools tend to get new technology tends to be adopted in silos. A lot of people say, that’s our CIO’s job, that’s our CTO’s job, et cetera. And I think with AI to really unlock the full potential, the economies of scale, et cetera, you really need an enterprise-wide strategy to avoid all of the takeaways we just talked about, all the potential pitfalls, all the issues, et cetera. You need to bring a team together, a team of leaders or a team of project owners to really figure out and prioritize where your opportunities are and go after the highest value, lowest risk opportunities. And so what I would encourage anyone, any company watching this is make that a priority. Start the discussion. Make it a company-wide discussion, not just an IT project, not just a contact center project, whatever it might be. And avoid the pitfalls of having a disjointed strategy around your adoption of these technologies.

Paul Weiss, Bluewave:

It’s an organizational charter and you can backfill specific needs and use cases across the various places where the benefits may lie. I love that we’re ending on exactly that theme with what you just said, Tim, because what you just talked about was better together from a human perspective, organizations collaborating.

The African proverb, if you want to go fast, go alone. If you want to go far, go together. I think that applies here.

Breaking down, siloed decision making that causes paralysis or disjointed strategies or fragmentation and complexity and cost. Those are very real challenges in businesses, large and small across every vertical. The better together element of putting people and minds and heads working together, patient measured, but also thinking big picture so that every step an organization takes is in the direction it intends. That’s a great place to end.

I thank you all for your participation in our event. Thank you for everyone that has watched or will watch our conversation today. Again, Paul Weiss from Bluewave, and I’d like to close out our session.