Barrett Thompson: Hello everyone. My name is Barrett Thompson. I'm the General Manager of Commercial Excellence at Zilliant, and I'll be your host for our podcast. I'm joined today by Lou Simon, vice president of Uptima Elevate. Lou, welcome to B2B Reimagined.
Lou Simon: Thanks for having me Barrett.
Barrett Thompson: Hey, Lou, just before we get into our topic today, would you mind sharing something that people are generally surprised to learn about you?
Lou Simon: Interesting enough, I don't always come to work with my hAIr done. I have a hat on today as we're talking through this podcast, and I think most people that talk to me on a dAIly basis, I'm usually pretty professional in my appearance. But that, and I also enjoy the outdoors a ton, as you and I have talked about on many occasions.
I like being outside. I like doing the sunshine thing and being active with my family whenever I'm not working in technology.[00:01:00]
Barrett Thompson: Sounds like a great place to spend your time. Well, I'm looking forward to getting into these great topics today. We've been having an ongoing conversation about how companies might begin to leverage AI in different forms, and there's several out there.
And of course the goal of that is creating better outcomes and experiences for their customers, for their employees, throughout the revenue lifecycle. And I'd like to begin there and ask you to frame up for us, really, what is the revenue lifecycle? How should we be thinking about that?
Lou Simon: Yeah, that's a really good question.
And, coming from Uptima and the really smart people that I work with day in and day out, it can be defined differently for different customers. As you guys are aware, when you talk about pricing and you start talking about what's important from an attribution perspective on the back end, from products, it changes customer to customer.
But from our purview at Uptima, we're a bunch of business consultants that [00:02:00] do love, and we get into the detAIls of what the revenue lifecycle looks like, and that has anything to do with the revenue generation cycle, which is quoting, understanding what pricing looks like in that part of the process, selecting products, understanding volumes if you're in a manufacturing industry, understanding a level of services at that point of the conversation, running that through the pipe to contract creation.
Based on whatever it is that you're quoting at the front end of the process, generating those contracts. It could be entitlements, warranties, there could be different ways that we do subscription modeling, from a contractual perspective, moving through the process. And then on the opposite end of it is how you actually start to recognize that revenue, right?
Like the asset gets delivered, the service teams are on the ground doing work, the software is being purchased and utilized in the space and everything end to end through when it hits the general ledger for our [00:03:00] customers. That's what we consider kind of the revenue lifecycle. And that's a very high, broad brushstroke, I think, but it can get very detAIled.
It can get kind of in the weeds too. There's fulfillment processes. There's a bunch of things that happen in the ERP, but for the base level, I think that's a pretty good idea of what we talk about typically with our customers.
Barrett Thompson: Well, we're fortunate to have you leading a breakout session at our customer conference called Mindshare coming up in a couple of weeks. Would you take a moment to preview what you're planning on talking about and who should attend your session?
Lou Simon: Yeah and thanks for having me by the way, Barrett. I've been blessed to really have these conversations with you guys. You guys are a great partner of ours, a lot of smart human beings over there on the Zilliant side of the fence.
So it's really kind of an honor to be invited to speak at Mindshare and what the conversation is going to be, and I'm hoping that the audience is a little bit engaging when [00:04:00] we get to that, is stitching technology together for industry innovation. And it's not innovation in its typical form when we think software innovation, writing new code, or coming up with that next great technology or really even kind of defining a new algorithm or anything in that way.
That's not the way that we're looking at it from an innovation perspective, it's really taking tools that we have at our disposable or that are disposable to us and packaging them in a way that it makes sense for our customers to really get the best bang for the buck out of all of these different platforms.
And Zilliant is a really good highlight to that fact because we are kind of looking at the everything to revenue model. We just went through the revenue lifecycle. Our customers needed the ability to be able to do that real-time pricing, be able to model that in, put the boundaries around sales teams, [00:05:00] around operational teams, around the rev op teams, and package that in a way that it's one agreed upon “this is actually how we do our pricing.”
Two, it's done in a fashion which is expedited from running many deal desks across the board or having pricing be two weeks old, three weeks old before it's even used in a process. It was a very important part to bring forward for our customers, right?
Like, so as a solutions integrator, systems integrator, we're always looking to make sure that we're mapping everything that we do to our customers’ needs and in the industry, whether it be manufacturing, high tech, med tech, or services organizations. That pricing, AI, bringing that to the forefront and then putting it on a platform like Salesforce is an extremely powerful way to really start thinking about how we stitch those technologies together and whenever we go front end to how Salesforce is implementing this now.
We do a lot of revenue cloud [00:06:00] implementation on Salesforce platform because we think it is one of the better tools in industry to be able to apply to the front end of the revenue lifecycle. The quoting, the contracting that's avAIlable now, in the tool, being able to shove that into what's called a sales agreement from an industry's cloud perspective is extremely powerful for people in manufacturing or in those med tech businesses to run their run rate business.
So it drives everything from quoting to pricing, to contracting, pushing that into an agreeable forecasted revenue stream that then can be ordered and managed downstream and show actuals versus forecasts. So now you actually can get into operational movements within companies that maybe it wasn't even avAIlable to them before we were able to package these technologies together.
So it's a really powerful concept. I'm passionate about it. As you know, we've had a lot of conversations around doing this, [00:07:00] and I'm really excited to have that conversation.
Barrett Thompson: Well, that’s going to be a great session and indeed you've been in a position in your career and I see it from Zilliant side as well, that process innovation that you spoke about inside the business, innovating customers already think about how to innovate in the products, what they sell, but they may not be investing as much in how they sell. And that's what I heard you describe there, the end-to-end revenue process, technology enablement coming along to innovate within that process is a powerful combination. Lou, the mention of AI there, I think this is something that's top of mind for many people that I've been speaking with for years.
In fact, maybe with some very new energy coming into it recently. You know, what is your perspective really? I'm thinking about that two ways. What is the promise of AI? And [00:08:00] then let's be pragmatic for a moment and ask whether or not that promise is being fulfilled or what the gaps might be to fulfillment.
What are your thoughts on that?
Lou Simon: Yeah, so these are fun times when we're discussing new technologies and you and I, both Barrett, have really been in a really interesting time in human evolution where we've had these groundbreaking technologies hit us several times in our lifespan, right? We had the internet, we had web 1.0, web 2.0, mobile phones, right?
Devices in our hands that do more computing than sending astronauts to the moon in the seventies and eighties, right to Web 2.0 to the service level architectures, the streaming media that we're able to do now, networks and cell phone towers that can actually give you internet all over the world.
There has been so much technological advancement and now we get to see this latest one. [00:09:00] And I just got goosebumps thinking about what the capabilities of AI might give us as human beings. And everybody's really thinking about GPT, like these generative models and all that fun stuff, and they're really in their infancy.
Right there, there are these tools that really can supplement what you do on a daily basis. They can give you some ideas for content. They can give you ideas on how to package verbiage. They can summarize briefs. I use a lot of the GTP models now to read papers and summarize it for me so that I can actually see the content and then go back and read if I need to.
I think there is a large application moving forward with that. But let's take a step back. AI has been around for a long time. It's not a new thing that just happened to be released in November when GPT was released. It’s things you have to touch on, [00:10:00] right? Yeah. Yeah. You know this right, Barrett.
Barrett Thompson: I do. I mean, I have to confess, I followed it pretty closely when IBM's Deep Blue beat Gary Kasparov at chess, right? One of those sentinel moments in the development of AI. But of course, that's a very limited domain that took many years. That computer was trained to do one thing and do one thing only, and it did it well.
So we've come a long way since then. But you're right, this is not the first AI breakthrough that sort of captures the imagination. Right.
Lou Simon: And it's, it's one of those things I think people lose sight of those. We've been using AI in our daily lives, probably for the last 10 to 15 years, like we've had it built into Google search, we've had it built into Netflix recommendations.
We've had it built into Siri in the palm of our hands, right. Google Assistant, all of these things like we have. We have AI. It is all over. It's all around you. You probably don't even know that you're being influenced by it daily at this point. Right. It's been behind the scenes, but those are all [00:11:00] directive logical Ais that are built to really drive to a conclusion.
The difference, and this is what gets me excited, the generative models that are coming out now actually can generate ideas. They can, they can see a large set of information. And present a number of different ways that you can approach that information.
Whereas before, with machine learning and robotic vision and a lot of these other AIs, they're built, directed for one thing and one thing only. And that's to complete that specific task or to yes, drive that specific set of facts. Zilliant is one of these, right? Like, you guys have a wonderful AI model that actually goes through, it's deriving facts, driving to a price recommendation, and it's doing that based on inputs that you feed it.
Yeah. So it's impactful and very dynamic in what it does, but it does one thing extremely well. It drives that price, it shows you what it should be today and gets your customers [00:12:00] moving forward with what they need to do instead of dealing with the pricing negotiations. So that piece of it is very exciting.
The generative stuff can go anywhere. And I just watched, or I was having a conversation here in, I'm in the city of Pittsburgh, and I used to work in robotics, so I am very familiar with the robotic vision and all of these PhD level people doing robotics now. Great. Hard work. I just want to tell you, it's a lot of work.
I was talking to a CEO of a company where they're actually using AI and one of the, if you have seen these Boston Dynamics mules.
Barrett Thompson: I think I've seen one that looks, it looks like a four-legged animal, right? It can chop up.
Lou Simon: Yeah. I think they're called Mule or something like that.
But this was a company that you wouldn't expect to be using this. And I'm not going to throw their [00:13:00] name out there as we're talking about it, but they were using AI in the robotic dog to service equipment that couldn't be serviced or could not be constantly monitored by people.
And that is a real use case. So these kind of concepts coming out and thinking about where technology is going, I am extremely excited about it. I see it as kind of an exoskeleton for our brains. But we have to be able to evolve with it. And a lot of the things that you're seeing now with prompt engineering and being able to talk with the machine, extremely futuristic concepts that we're seeing in our lifetime.
Barrett Thompson: So, you know, you and I, let's face it, we geek out on this, right? We're technologists and we've grown up in this. We're so close to this. What are you seeing from customers, from business leaders, right? Not staring at this as much as we are. Is there a mild interest? Is there strong appetite?
Like [00:14:00] what's the sentiment out there right now? Especially thinking about their revenue life cycle?
Lou Simon: Yeah. And everybody, so I think everybody is curious about it. And there's not like an individual or a direction yet on how specifically the generative stuff is going to be used. Everybody is thinking about it because we get asked. I have had an ask at least twice a week for the last three or four weeks, and it's getting more and more people are starting to play with it themselves. I have. As we're having this podcast right now, I have Auto GPT running tasks for me, because I'm trying to figure out a couple things on how it's going to present information like running in the background.
But to the point where we are, we're coming up with a perspective on AI and it's not dissimilar to what I said, like we have implemented AI at Uptima in the form of next best action, natural language processing, like to be able to kind of understand in the service side. We're [00:15:00] very good at whenever we have to manage those assets and there's calls in to a field service center or some sort of call center to be able to manage those call routings and do that. There's a lot of AI that goes into that as far as the routing, responding with a chatbot, those kind of things. So those AI built into those processes too. So, from an Uptima perspective, we've done these things.
It's not that we haven't been working in AI or haven't done these technologies. The generative piece is new. It's interesting, and it's one of those things where we're trying to come up with perspectives on a customer. There's a lot of different ways we can go with it and each customer again, is different.
Whenever you have the ability to formulate new ideas off the information you have, it takes a customer that's pretty mature in their data. In the ways that they've been able to plan out the information they want out of that data and then be able to [00:16:00] apply that, or custom grow, a large language model for that prospective customer.
It's hard, but I think it's going to come together more quickly now that we have some of these GPT tools.
Barrett Thompson: I'm looking forward to what the future will unfold for us. Do you have some thoughts? When I think about pricing and sales, you know, maybe toward the front end of that revenue cycle, are there some specific classes or problems that you think the generative AI technologies might be suited to serve?
And these could just be very speculative, you know, we're not going to come back in a year and see how close you were to accurate Lou, because no one can know.
Lou Simon: You're scaring me now. Now we're stamping it in history.
Barrett Thompson: But maybe just to provide some what-ifs or maybes or don't tell yourself no just yet.
Give the technology a chance. It might have an opportunity here, here, or here. [00:17:00] Do you have any thoughts on that?
Lou Simon: I think the world is our oyster at this point, and I mean that wholeheartedly when we apply a different perspective that's not human to a set of information and let's talk about end-to-end lifecycle of, there's so many data points across that it's incomprehensible for a human being to even really get into that information and predict patterns in the way that neural networks can, and generative AI can potentially do that.
So that's one piece when we can apply that in a way and come up with different ideas that we could never have thought about as human beings, that is one. Bringing it home to a more realistic perspective. When we think about the front end of a revenue life cycle, it's mainly driven by product and pricing, right?
Like, so we pick products, we price them out, we do a quote, we have a contract. we have these basic fundamental pieces [00:18:00] and being able to use generative AI to maybe say if you package this individual product, it could be something as small as a bolt with this particular item, you gain more revenue on the backend, more customer satisfaction on the backend.
You have the ability to do that. So start packaging this specific product with this specific product, maybe if it's somebody on the service side that's better. After you start servicing an asset on the ground, that could be a predictive model that goes back the other way. So we're thinking about these ideas and I think you and I had the conversation.
It's kind of like how McDonald's packages their value meals together. They tie a Coke on it. It makes it easy for you to buy. That Coke doesn't cost McDonald's anything, but it increases the price of the value meal. It's called a value meal, they're making more money off of it.
It's easier for you to purchase it. And that's all [00:19:00] driven by some of these things. So those concepts, again, McDonald's does these things. They have AI that runs in the background to learn, but generative AI is going to be able to predict different ways to sell these in ways that we didn't even imagine.
And those kind of things are what I'm looking to start to uncover.
Barrett Thompson: Well, I'm excited by that too, in the sense that if I can characterize what you're saying, AI may see things that we miss, right? It might actually find some insights that we're not finding ourselves. And I think that's exciting.
We certainly saw that happen in some other domains. So, we mentioned chess a minute ago. The purpose built Deep Blue just a few years back. Did you follow this story? When Alpha Go, a part of Google's Deep Mind built a machine to play the world's most popular board game, Go. Go. And there was a five game match against South Korean [00:20:00] professional and World Champion Lee Sedol.
And the machine won. Which I think shocked a lot of people. This is a lot more subtle. It's not as deterministic, you know, as and, and even since that time, well, when it played, what was discovered is the AI was making moves that no human master on the planet would ever have made. And they weren't. At first they thought it's a mistake, right?
That's a very subtle but extremely powerful move. Right? So finding the hidden gem that we didn't see, I think, you know, that'll get the headline, won't it?
But there's a part of me, Lou, it feels like maybe the bigger value add on a day in and day out basis is not that the machine is necessarily superior to what the best human on the planet is doing, but if it could even measure up to what an average or [00:21:00] reasonably good person is doing and just do it at a scale that we've never seen before. You know, go study a thousand customers in my customer database and figure out things like what products I ought to be selling to them, which, if I had enough time, if I had infinite time as a human, I could probably go build some Excel charts and pull it up and see the magic myself, but I don't have the time.
Lou Simon: You couldn't, not even if you had the time, you'd get bored, like human beings get bored. So AI does not get bored. It is ruthless. So these kind of things, this is what is really kind of exciting to me and why I think it will begin to eat the software world is because it is ruthless in nature. If it gets a goal, it can figure it out.
It continues to turn away to get at that goal. And to your point, it will create ways to do things that we never thought even possible from a human perspective. And that's because it is just looking at it from [00:22:00] a different perspective. That doesn't mean it's smarter than us. It doesn't mean that anything other than it sees different ways that it can connect the pathways that we don't do as human beings.
I will tell you, it's about as smart as a six month old puppy at this point, as far as the auto GPT in it, trying to find things. I know there's like purpose-built AIs out there that do much smarter things in a finite manner. But when we start talking about this generative stuff, I've asked it some interesting things and it just kind of recites stuff back.
So I equate it to my puppy, whenever I'm trying to train it. But I think finding those high value problems and us as people thinking about ways that we can apply the technology to it to get to elicit the response and the solutions is going to be just as important as the technology itself. So I think it really is in the hands of who's using it.
It's a lot of power to wield.
Barrett Thompson: Lou, are there pitfalls that you [00:23:00] think companies should avoid as they start exploring things like the GPT technologies?
Lou Simon: Yeah, there's a few. One, it's not as easy as it appears whenever you just start typing into chat GPT. When you're doing it from a business perspective, there's data proprietary concerns.
There are data management concerns, the model concerns. Is it giving you the right information? Who trained it? These are really important pieces to think about whenever you're going at it from a business perspective. And again, when we're doing this, bringing it back to the revenue lifecycle, if you're talking to a systems integrator or a consulting group that doesn't know financial data and doesn't understand how to consult across the revenue life cycle and bring that back to how you train this AI to get the right responses that you want off of your pricing data and you just apply something blindly or apply it with a perspective from a software engineering group [00:24:00] you're going to get a different model and the model if you're using that to kind of condense information or look at other ideas for your business or predict contract pricing, for example.
It might not be correct. So there's a lot of thought that must go into it and right now, like my biggest thing, especially with these open ones, don't just put proprietary data into it. We've, we've had a couple stories along the way of engineers kind of playing around and dumping very important intellectual property into a very public open AI, so make sure that you're using in the right way to make sure that your data is private and you're not losing information.
Barrett Thompson: Yeah, that's critically important. Your point about validation resonates with me. I mean, even in the very strict narrow purpose-built models that [00:25:00] we build and deliver at Zilliant, we must go through validation steps each time we deploy that on a certain customer's data, and there's at least a couple of reasons.
There're knobs and dials in there that we can turn, you know, parameters that control how the algorithm behaves and businesses have different goals. They have what it means to get a good answer out, can be contextual. It can be different from business to business. That's a part of the validation and tuning as well.
And I'm sort of imagining that that's at least as true, perhaps even more true in some of these generative models. So I take that recommendation to heart - validate what you're getting out before you just take a mechanical answer and run with it.
Lou Simon: And again, my background's in high tech software and I've done robotic vision, right? So software in these places, it's always validation. And from a robot perspective, it always has to localize where it's at to figure out where it's at in the space. [00:26:00] Even if two robots are built the exact same way, with the same components, with the same algorithm, with the same idea of where they localize.
There could be a calibration or one could have a flat tire that takes them off of a course, and it's not exactly as the same as the other one. So it's very important to think about what information that you want out of it, how important the front end of the system is, and that's where I think if you bring someone in like Uptima that really understands the revenue lifecycle, understands products, attribution, what you're trying to get out of this and brings that consultative advice to the table with our customers day in and day out. We have these industry perspectives now, from our Elevate brand that it's because of our deep thinking around these problems and revenue that we're able to come to the table with a lot of that context.
It's why bringing that consultancy to the table whenever you're thinking about this, is much more important in my [00:27:00] perspective than let's just apply some generative AI to this and look, we have AI. And a lot of people are focused on that very shiny, “we're doing AI,” but are you doing it right?
And that’s what I think it gets a little bit interesting.
Barrett Thompson: Well, Lou, it's been a great conversation today. I am looking forward to what the future holds. I know your customers are in good hands. I think ours are in good hands and together we can do great things for them with AI and do that prudently and to great advantage.
Is there anything else you'd like to share with our audience today?
Lou Simon: No, I think this was great. Thank you. Thank you Barrett. It's been a wonderful time getting to know everybody on the Zilliant team and having our time outside of the professional sphere and really getting to know one another.
And I hope to be able to do that whenever we're at Mindshare with some of the customers on the ground and really shake some hands and I hope people listen to this podcast. And feel [00:28:00] comfortable enough to come and have these conversations because it is really what I like to do. It's a never-ending curiosity, and that's why I'm a technologist.
So I invite that. Please come and talk with me, talk with the Uptima team. We're all in the same boat.
Barrett Thompson: Your passion is contagious, Lou, and we look forward to seeing you in a couple of weeks in Austin. I want to thank each of our podcast listeners for being with us today. In the show notes, we have a link to register for Mindshare.
Yes, there is still time, and you can hear more from Lou and many other Zilliant partners and customers about how they're powering intelligent commerce. We are committed to your success, and if you need any assistance, please reach out to us at Zilliant.com. Would you do me a favor and rate and review the show in your podcast app today as it helps us to continue to put out great free content?
Until next time, have a great day.