Think Fast: The Agentic AI Roundtable

Hello everyone. Welcome to this webinar. It is excellent to be here with all of you, and I'm so excited to share some of these things that we've got. All of us have heard a lot about agentic AI; there's a lot of hype, a lot of excitement, and hopefully today we've got some really interesting research, some really interesting customer stories to share with you, help you get a little details on how you can start today to get more value out of AI, more value out of your data.

And joining me today is an awesome cast of characters. I'm Jason from Domo. I've also brought along with me, Miles from Dresner, and Bruce from Torchy's Tacos. And we're gonna share a few things, have a conversation. Why don't you both give a little wave, say hi. Awesome.

Well, to start us off, the Dresner team has been doing some amazing research looking into how companies can take advantage of AI, what they're doing in Agentic AI, where they're seeing success, where they're seeing challenges. And we wanted to bring them today and Miles specifically to talk a little bit about the research that he's doing and tell us how it applies to our own businesses. So, Miles, take it away.

Miles: So I'm really excited to be here to share our research on the state of Angie AI. As has been suggested here by Jason, this is an area that's gotten a lot of market attention. And what we've seen is not just vendors, but organizations are really jumping on this. And I'm gonna share with you some of the latest research we have, which is really startling when you look at the numbers in terms of adoption. And I'll tell you where we were six months ago and where we are today.

First a little bit about Dresner. Dresner is in its 18th year as an advisory company, in market research company. We think of ourselves as being unique because we're more business oriented. What are the outcomes? What does this technology mean? Where is it going, and how is it going to impact you?

So this is the latest research and the most interesting thing to me is the number of organizations that claim themselves to be in production right now. It's 15%. Now, to give you a sense, when we did this roughly five to six months ago, the number was six and a half percent. So that's a pretty startling increase. And we've had a doubling of organizations who are actively, you know, experimenting with this technology. So to see within that kind of a period adoption and experimenting jump at that level is pretty amazing. I cannot think of very many technologies that have seen that kind of acceleration and impact.

And, you know, it's interesting. There was a venture capitalist I talked to about six months ago who suggested to me that in this whole move to generative AI, that this was the real pony, the thing that was really going to change markets. And so we're starting to see that now in terms of perspectives, which is a little different measure than just adoption. It's trying to look at where are people in their journey, how do they perceive this? And what we see is a pretty amazing group of active adopters running at about 28% right now. So this is mixing together the folks who've already adopted with those who are gonna be in that early adopter, if you think of it in the Jeffrey Moore way of approaching markets and adoption. So that's a really healthy number.

And then we've got this next group, which claims that they're gonna also be an early adopter. Those numbers are enormous. They're not the way typical markets evolve. So yes, there's a lot of hype here, and yes, there are concerns about, you know, how successful people are gonna be and how fast, but the fact is the numbers look really good. And what I'm gonna do in just a few moments is to say, well, what are the characteristics of this early adopter group? And what things do organizations need to have in order to be successful?

One other thing that's kind of interesting when you look at it is where is this adoption happening in the business? And it's interesting because, you know, research and, and sales and marketing and those kinds of organizations are obviously part of the early adopter group. They're leading the charge forward finance, which you're gonna hear about in just a little bit, is, you know, a little bit slower. But in talking to industry leaders recently for an article will come out in December, what we're seeing is people view finance as being revolutionized here. This may be as big as the impacts of the spreadsheet. And for those who don't remember, the way people used to do financial analysis, as Bruce can attest to, was with a piece of paper and a calculator. So this may be as big of a transformation and hopefully Bruce can talk a little bit more about what he's seeing as an actual finance leader.

Another thing is that we, you know, obviously see companies that are ready for new technologies to be early adopters. So obviously technology organizations and business service organizations have received a lot of attention. It's been interesting, some leaders of business services organizations have indicated that their teams either need to become AI savvy or they need to find another profession. So we're seeing a lot of adoption there. And that doesn't surprise us, given where the CEOs of companies are approaching things. And in fact, I did an interview in the last couple weeks with a company where they basically said, we're AI first in everything we do. So we're in the midst of seeing that, and the industries are showing that off, in particular in business services.

Now, it's very interesting. You know, for a long time there have been real consequences to businesses who have been successful with business intelligence. They have maturity in their data processes, they've industrialized, as my friend Stephanie Warner likes to say at MIT scissor, their data. And we're actually seeing that in the data. So the ones who've been early adopters who are in production have tended to be in organizations that have three characteristics, and we're using these as proxies for industrialization. So the first is they've been completely successful at their BI implementations, or at least somewhat successful. It's really interesting. And a lot of those people who are just experimenting are also in this group that are somewhat successful or unsuccessful.

Another interesting characteristic, which we of these folks that we find is those who've been completely successful with BI tend to have a broader view of how to use this technology. They don't just look at it as, Hey, here's a way to cut labor. They look at it as a way, here's where we can transform customer experience or value we provide customers.

Another very interesting thing is success with self-service. This is, you know, a really important thing. And, and the reason I think it's an is that how do you get successful with, either in terms of one angle is, you know, better getting data to the user. But the other angle is applying this to business processes. How do you get successful if people can't find the data and they can't be able to make use of that data quickly? And so this is another really good proxy in terms of what organizations who are successful seem to have as an experience step.

In terms of priorities, obviously everybody is looking at the productivity efficiency side, but most of the organizations that are really smart about this, I will add, are looking at it, how do they augment what people do? And in that augmentation, how do they help in other areas as well? So decision making, customer experience, finding the information inside the organization. But what we think people are gonna do over time is redeploy labor to expand the business. And the organizations who are early adopters are gonna be the ones who are gonna be able to make that jump from just getting better efficiency in these organization to actually, how do I grow my business? How do I create new value for my customers?

Obviously there are obstacles. You know, this technology is built on gen AI in terms of how it connects and shares and, and turns, you know, data and process into action. So obviously it's very smart for organizations to think about these things early and make sure their designs sufficiently address all of these issues. Obviously quality and those kinds of things in terms of data are critical. That comes back to that whole industrializing process because if you've industrialized your process, you have accuracy. All of these other things, you know, really are built on. One other factor I haven't shared, which is leadership, leadership is critical to making the business case of the business invests in this capability and aligning other initiatives to mature processes. My my recommendation to companies who aren't as data mature, even if they're in silos and spaghetti, is find some early wins, but get the business now that it's excited about this to invest quickly and getting your technology, your bases, you're industrialized.

So just some conclusions that we are seeing as we look at this marketplace, that are relevant. The first is we're seeing this erase boundaries. There are a lot of different areas within the organization that had very specific places that they sit. And we're seeing software organizations, you know, all kind of go after this. So the organizations that are strongest in the traditional capabilities and doing that maturity thing are, are, are going to be early winners here. Our research from last year, I don't have the numbers yet for this year, shows 68.5% of software companies all have some kind of offering. Now, to be fair, they're all maturing. They're adding capabilities that are essential. They've initially put something out there 'cause they felt they had to have something out there. So we're seeing, you know, the best companies have made it easy for you to go and implement these capabilities.

Early identifier and enterprise adopters have what we've been saying all along, mature data capabilities, strong leadership and mature BI capabilities. Adoption is surging and rapidly approaching, approaching generative AI, which will be more on the personal productivity side, as agents become one more prolific. Here are the opportunities. We've kind of talked about that. And then one other thing you know is really important is training. You don't just hand, you know, a platform over to somebody and say, go run with it. And one other thing I personally think is gonna be critical as time goes on, we really want to move this to the domain people. They should be the one that are tuning and training agents over time. And so the smart organizations are gonna be those who enable that to happen.

So there you go. I look forward to sharing more in a few minutes.

Jason: Awesome. Miles, thank you. I wanted to see, I love something that you said there about the top three. I call 'em like the keys to success with agentic AI, right? I wanna make sure that nobody misses those, the things that I heard you say, BI implementation, self-service, right? Making self-service BI and making it that easy, right? Which I think has a lot to do with the quality and the organization of the data. And then the third one I heard was data leadership and making sure that that is in place that you're leading from the front. Any other thoughts on that? Miles? Have I captured it accurately?

Miles: No, I think you have. I mean, the only other thing we've kind of seen is some ability, you know, some experience using machine learning and data science and those kinds of things are is important. You know, those people just kind of, they're more comfortable. They can explain to the business what's going on. You know, those that are immature, you know, this is a whole language that they have to develop. So, no, I think those, those three things are absolutely critical and, and the organizations that are there are gonna take advantage of it. And, you know, there were some numbers that that, um, MIT had a few years ago in terms of profitability and other things like that. And those numbers were good, when you looked at the difference in terms of means for different industry. But I, we think that it's gonna be even stronger because if you really can deploy this, the efficiencies you're gonna gain are gonna be dramatic. The transformation you're gonna be lead is gonna be dramatic. And so it's, it, it's an, it's an opportunity for those who are there to take advantage of those who aren't there. And so it's time for you to get going if you're not ready, if you're in silos and spaghetti, get going now.

Jason: Love it. Thanks Miles. And I'm sure we'll have some more questions for you here at the end. I love what you're saying about these steps and these processes and, you know, we're about to bring on Bruce, who's literally lived this life of being a data leader and bringing in and doing this with agentic AI. Before we get there, if you've got any questions on the chat, go ahead and drop those in and we'll see if we can't answer those during the session.

But I wanted to share just two quick things that we're seeing on the Domo side before we hand it over to Bruce and get his experience from the field of doing this as a business leader. So this is the report that miles has been referencing, and this is the vendor portion where we're looking at how the different companies and vendors in the AG agentic AI report for data, AI and analytics platforms performed. And I get the great privilege to represent Domo as a leader and a, in the three-way tie for first place in this report for agentic platforms.

I think a lot of that has to do with what we've done historically. And it goes back to everything that you're talking about. Miles. Organizations exist to organize resources to solve a problem. And that has been difficult. And that's always been our goal at Domo, is how do we help people take their data and use it to drive business outcomes and solve their problems. And so we've historically provided all of those tools that you need to unlock the data that's in your business, right? Whether that's connecting to external data, visualizing it, exploring it, helping make it accessible to the business, turning that into alerts or insights, and then using it to drive workflows and actions and external systems.

And a lot of that is very task oriented, but as what we're seeing inside of Domo as well is that with AI we're becoming more goal oriented and helping people make that transition from having the data and using it to think about different types of analysis or do different types of analysis, and really moving that into a place where they're driving outcomes and they're trying to use that to optimize their business, turn it into results, which is really, really exciting to see and really cool.

But I think, like you mentioned, two of those three keys to success have to do with having good BI foundations, having more self-service and more accessible data. Both of which I think makes so much sense with what Domos tried to do. Again, it's an extension of what we've always done.

So anyway, one of the things that we've seen success with our customers is when they think about this as 70% of the value that they get from AI is from just incrementing, right? Improving things that they're doing, and doing them better, and then improve reinventing what people are doing, doing it in different ways, thinking about different ways that AI can help them do those jobs. And then only 10% in innovation. So it's interesting to think of how transformative AI has been. You know, I love Miles you mentioned. It's could be as big as the impact of the spreadsheet, right? And I think that now we see spreadsheets maybe not the sexiest thing, and yet how much of an impact it has on people's ability to organize their data and just orders of magnitude more efficient than, you know, writing that down on a piece of paper or trying to remember it all in your head and how AI can really move you in that direction. But it's gonna become about how do I increment what I'm already doing, how I improve the things I'm doing? And then finally, how do I innovate, do new things I've never done before?

Which brings me to our next guest, Bruce, who is a spreadsheet wizard I am sure in a prior life, but is not afraid to get in and try something new, improve his processes, and do agentic AI. Bruce, four times Domo Sapien at this point, bringing Domo and implementing Domo to company. Tell us about your journey. Tell us what you're doing, what you're excited about, and what's new at Torchy's.

Bruce: Yeah, so we're doing a lot of really cool things, and I really did like the C criteria for success for AI because it is, that success with BI is just rings so true because in so many cases that bi that foundational data, that highly enriched detailed data, that's what powers AI. So without that, your AI doesn't have that power, doesn't have that data, doesn't have that intelligence to do anything. So it's like, it absolutely, that's, they captured it perfectly on that.

So, let me show you one of my favorite slides. So this really to me just absolutely encapsulates everything we're doing all in one slide. This is our strategy. It's a multi-tiered strategy. But it is, it's people, processes and technology. That operational efficiency, it's turning that operational environment data into intelligence. And it's moving from descriptive in that bottom right, that descriptive analytics, that's what happened in the past. You need that, but that should be served up automatically. And you wanna move up the value chain to predictive and prescriptive analytics. And that's what you wanna move the predictive being machine learning, the prescriptive being the AI.

So we have a multi-tiered strategy, using AI to bridge the gaps because people speak different languages. Stores and accounting speak different languages, have AI bridge that gap by enriching those data sets. And using AI like JI AI workflow to optimize operational processes. And then using the AI and machine learning to get deeper insight as to what's going on with the business, such as with customer reviews and other areas. And it's just demos allowing into that, with focusing on that architecture and system design that no-code, which is so powerful, important. And that's where technology's moving to with that no-code environment allows for quick delivery. So that is our strategy that really encapsulates really everything we're doing.

And then I'll show you quickly a couple things real fast too, of what we've actually are working on and put into play. So this is an example of an indicate I workflow. This is a vendor approval process. So this leverages the AI, leverages that data, and allows us to improve processes. There is that wonderful AI instructions, I want you to do this, I want you to do this using, once again using that no code, using that real world speak. And that is what just we're telling the AI what to do and how to do it in a no-code way, which is fantastic. And then what that leads into is applications like this of that vendor process of being able to request bills, being able to have that insight all served up using AI, using the platform to improve those operational processes, make them more efficient, reduce errors, and make everything just operate much faster. So, those are a couple examples right there of what we've been doing. So, at a high level.

Jason: Awesome. Bruce, I love that idea of going from descriptive to predictive, and how you're trying to help these people at your business, right? The accountants, the finance people, or even the people there running the stores move from, you know, doing something that is necessary but is maybe lower value and moving them into something that is so much higher value. Tell us a little bit, Bruce, about, you know, some of the specifics of how you, how you started, like what was the first thing that you tried in agentic AI to get your business moving in that direction?

Bruce: Yeah, so we have a lot of manual operational processes in the finance and accounting area. Very just heavy keen in data, things like that. So, knowing how the, how Domo and AI combined can take input that data, and leverage it against our existing highly rich detailed data sets, which that's where all the data comes in. It first starts, it comes in, we import it using the connectors, we enrich it. So we have high, high detail and then we enrich it, which is, which is exactly what, what the AI machine learning needs. And seeing those processes that were the ripe for automation, understanding the process, but then understanding the underlying data to put all those together to see what makes sense, what we could execute against, what were some major pain points. And and then extending the ERP, so that extension too. I like how we're extending other applications in this case ERP, some limitations in the ERP of it, not having AI, not having a no-code environment. So being able to extend the ERP extend that functionality. So functionality that's just, it's really not good at, that combined with the AI, that combined with that data, I'll putting all those three together to really increase those, the efficiency, those operational processes, get things more accurate, faster, and better. So we can focus on, once again, moving up that value chain and, and move everybody here up the value chain. 'cause it does make people nervous. 'cause we have seen all the companies and go, I'm just gonna lay off people because of AI and, and, and they're doing this before, before maybe they even have ideas. It's like, can we not get the ideas first and then see if it makes sense? And then maybe if it leads to that. But I, I like to take the approach more of we're going to improve your skills, we're gonna move you up the value chain. This is not a personnel replacement. This is replacing processes, making process more efficient so everybody can win. Torches can win. The, the team can win, they can learn more advanced functionality and more value chains and, and look for more places for revenue and cut expenses.

Jason: So the question I have for you, Bruce, is what do you see as the future of the finance professional, right? What does their job look like a year from now, six months from now, a year from now, two years from now, versus what it's looked like in the past? And how is AI helping you get there?

Bruce: Yeah, so how I see it evolving is the very manual processes. I have to key in all this data. I have to update this data, I have to look and see what my accruals are gonna be, things like that. And spend so much time of looking at past data to generate this that we can use the, the Domo platform. We can use the AI and ML to do that because a lot of stuff predictive. So replace those processes and then having them look for places we can, we can revenue opportunities, places we can cut costs more savings and things like that. So that's more value to the company because just all the manual data entry, all that keying in is just, there's, there's a lot of people that can do that. But if, if we can replace that, then we're, you're gonna add much more value to the company to help us save money, to help us increase our revenue. And that's really, really valuable. Everybody wins. Plus at the same time, they learn, they understand these automation, this automation processes, the AI, the machine learning, they, they see how it works. So now it spun, spins up ideas. So once they see how it works, and that's what's happened a lot too, once they see how it works, go, oh, wait a minute. So that's how it works. That's what it means. That's machine learning, that's AI, this is predictive, this is prescriptive. Once they understand those concepts, then it's great to see the wheels turning and they go, oh wait, so if we can do that, can we do this? And can we do this? And can we do this? And with these tools and with these technologies, I, I almost never say no, it's almost anything is possible. That's what's wonderful too, because it is the art of the possible, these technologies are allowing that. So we can pretty much have the world at our fingertips to be able to do almost anything we want.

Jason: Bruce, you mentioned earlier about AI bridging the gap between, you know, the finance team and the front of house, the folks there in the restaurants. Tell us a little bit about how you're using Domo and AI to help bridge that gap and kinda the communication gap right around data, because I think that really relates to what Miles mentioned about making sure that you've got self-service as a core piece of making this successful.

Bruce: Yeah, so one, one great thing we're working on right now is, you know, they're the stores and we have the monthly accounting cycle. We have the profitability, we have expenses, things like that, that gets posted to our general ledger. But the stores, they don't have direct access to the general ledger. They like to see what's going on, but they also speak different language. Somebody will call something Ritz, somebody else might call something triple net. So they're speaking very different languages. And classic in the past is they just, it's just this miss and people just don't understand each other. So it is using that financial data, enriching that financial data and using things like AI chat, so people can ask questions in real language, how they speak, they can ask it completely differently. And the AI has learned our data, we've enriched it. So the AI can understand, I understand what this person's asking, and I understand what this person's asking. So it can answer different questions the same way, because it knows both sides, because it's been enrich, it knows our data. And that's, that's a great use of AI as to having everybody speak AI like the translator. It's almost like the Rosetta stone of, of, of it knows what everybody's saying. And it gives everybody the right answers regardless of how you ask the question.

Jason: I love that Bruce. The Rosetta Stone AI is the Rosetta Stone. I have definitely seen that personally in our work, and it is very true. Even looking at our sales organization, our marketing organization, our product organization even inside of Domo, they use different terms often to mean the same things. And I can only imagine, you know, your finance folks and then the people there in the restaurants and how they talk about different things. Definitely cool to see how AI is helping one source of truth dataset be usable by so many more people now because now it's accessible to them and how they want to talk about it. Really transformational, like the kind of thing that you can't even think about doing something like that without so much additional work. And, and duplication of data and just all the ways that people tried to solve that problem historically. And now it's, you know, you just open up AI readiness in Domo and add some synonyms and right. And have AI start going down this road of making it better for people. Very cool.

Bruce: Yeah. Another cool example is, so we have all our customer reviews, we have all the customer text, and one, one really powerful tool in the AI is being able to have the AI analyze that data. But the great thing is, I can create a custom scoring rubric. So a store number one can say, I want you to look at all my customer reviews, and I want you to score me based on quality of food and employees friendliness. Score me on one to a hundred, gimme the suggestions, gimme the complaints and make recommendations. But store number two, they're like, well, our quality food is good. I think our friendliest employee is really great. We wanna see speed of service, how quickly we're delivering the food, and do people know about Torchy's? So then they can ask different questions, and the AI will look at that and score it on that set of questions on that scoring rubric so everybody can, can find out, get deep insight as to how their store is doing, basically with bespoke questions, with a bespoke analysis that AI does. And it can score in multiple different ways. And that's, I think another fantastic use. Looking for problem areas, my problem areas, which may be different from his, but it's going to answer both. It's gonna score both. It's gonna gimme output complaints, it's gonna be recommendations, no matter what question I ask.

Jason: And that's another really cool use that's really Leveraging AI at scale, right? It's, it's one thing to have, like my prompts that I use, it's another to be able to share those with the organization and let people tune what they're asking AI, but providing enough guardrails, right? That it's accurate or that it's consistent in how we analyze that data. That's pretty interesting. Bruce, I know I've heard a lot from folks talking about AI, that AI is only as good as the data that we put into it, right? They can't, otherwise it's hallucinating, it's making things up, right? So we've gotta have good data there. Tell us a little bit about how you're getting the data and the amount of data that you're using to put into these and, you know, the challenges or what's, what you're seeing success with as it comes to getting the data and the AI into the same place so that you can do that kind of analysis.

Bruce: Yeah. Well, the great thing is having, having your, your BI, your analytics, and your AI and the same tool's, incredibly powerful because it starts with all the connectors. I'm bringing from, from our ERP, I'm bringing every single transaction line, every single journey entry line, every single invoice line, from our point of sale, I'm bringing every single taco or queso or beer that we sold, from our customer views, I'm bringing every customer comment in. So that's where it needs to start. You don't want that highly aggregated, summarized data because that's not what the AI needs. That's not what the machine learning needs. It needs that detailed data. So starts with that, and then I enrich it. So then it's enriched with store formation, with demographic information, other things like that. So now the AI machine learning can not only see the core data, but then it understands the, the demographics around the data. So now it's, it's just like a person. It's going to start understanding it better. Like, okay, wait a while, I see these sales. Well, you want, you always wanna find out what's driving sales. And when you enrich that data, that's what you, when you start to see, it's like, okay, what's driving, what's hampering sales? Things like that. So creating those highly enriched detailed data sets is the foundation that are reconciled against. So they have to be correct too, so that I auto reconcile these, I have alerts, things like that. If there's any issues, I know right away. But that's the foundation of all of this. That's where it all starts. 'cause if you don't have that, you don't have the AI and machine learning on top of it, but Domo enables that. And then it, it's there and then it feeds right to demos, AI, machine learning engines, for the analysis. Very

Jason: Cool. Every taco, right? Every taco, every queso,

Bruce: That's correct. Every single taco beer, queso. Because that's what you need. If I just had one line that said this story did $10,000 in sales, that doesn't really tell me anything that's

Jason: Going back to descriptive, right? Instead of actually getting to something that you could say, oh, let's predict how this is gonna work and let's capture things. I love the alerting that you mentioned as well. It's about capturing those things as they're happening, not waiting and getting a quarterly report, you know, doing a period close and running into data that surprises you, right? But instead being on ping on the pulse as it's happening.

Bruce: Exactly. And like one of the examples of enriching too is, uh, a detailed sales data, but one of our stores is at LSU, it's literally right across the street from the football stadium. So putting the academic calendar because it's right across the street from a college and the sports calendar. So now putting that against the sales data and then having the machine learning, I learned that. And now you can see what's driving, because sales may increase six x on a day of a big football game. But then there's the natural ebb and flow of school is in session. There's spring break, there's graduation, there's big athletic events. And then putting all that together, then the AI and machine learning can see it's like, oh, I see this is the kind of the natural ebb and flow of people around that area. Here's the ebb and flow of our sales. And you put that together, could start learning. It's like, okay, this is what's driving sales. This is when you can expect big days, but then when this happens, you know, not as many sales.

Jason: So that's one example of that enriching that data with basically what's happening around the store. One more question, Bruce, and then I want to bring miles in and we'll talk a little bit about the future and what we see and, and get some ideas. You and I talked a little bit about manual data entry and some of the tasks that your teams used to be doing. You showed us that example with the W-9s. Tell me a little bit about that and kind of where it was and, and what you're seeing now, and just as a really simple example of how people could get started with AI.

Bruce: Yeah, so that one, that process was, we have many triggering events where we need to create vendor information in our ERP. And in the past, what was designed is you had to manually key everything in, and then it went through a manual process if a manual review, and then somebody had to upload it. And, and there may be issues because nothing was validated and everything was manual. So now it's leveraging the agent AI workflow of instead of keying it in, guess what? You can upload the W-9. So now you, you take out that, that potential for keying errors because it now just reads the W-9, and then it not only reads the W-9, but then it looks for different, there are different boxes on the W-9. So if you see these boxes checked, then that's one type of vendor versus another vendor. Also too, that all that data that is in Domo, we have all our existing vendor data. So then it can now check to see, Hey, does this vendor already exist? You know, if it does, it kicks it back out, says, Hey, guess what? We've already so checked against all your data done. Well that guy already exists, so we don't need to recreate it. And then if it's not there, it goes on its trail. One of the key things too we had to do is, this is an accounting process. This needs to be an auditable process. So I need to see, okay, who entered it? And then it, it's gonna go for approval. It goes for, you upload it and it goes for review first by the person who entered it. Hey, is it okay? So we got that checkpoint and then it goes to an approver. The approver can approve or not approve it. They can see everything they need. If find a accounting has to approve it for the entry into NetSuite. But all of those processes using the AI workflow and demo, really has helped that process of, of removing manual entry issues, checking to see if the data's already there, plus then we're recording all those review points, all those approval points, because now it also has to pass an audit auditor may come in and go, who, who approved this? When did they approve it? How did they approve it? And now we have the entire audit trail.

Jason: And you know, I love that idea of you're building on top of what you already have. You already have this data here. You already have, you know, with workflows and things inside of Domo, you already have the ability to bring in the humans, right? To loop in with the AI and, and really do those auditable activities, right? And, and supplement the human tasks with the AI tasks and, and capture those back and forth. How long, Bruce, right? I think that's what people are gonna be asking. Okay, that sounds cool. Maybe I'd buy something off the shelf. Or, oh, did I have to buy a whole software development agency to go and build this for me? How hard is it, Bruce to stand up? Something like that, you know? And obviously it evolves over time, but what's kind of the, what would you say to someone who says, I need to do that, I wanna use AI for that?

Bruce: Yeah, so, there's a little bit of a learning curve to some of the workflows because they are incredibly powerful. But, time to value too is another big one. Such as that. You saw that two sided. We have the vendor, we have the invoice, that, you know, you, you can, you can spin that up in a matter of two, two weeks. Whereas, whereas in the old school world, processes like that, using apps that are heavily coded, things like that can take 3, 6, 9 months. So your, your, your time to value too is increased significantly due to the, the tools there due to the concept of no code. And due to the AI as to how you tell AI, it's not coding. I'm not writing Python, I'm writing real world commands to go, I want you to do this, I want you to do that. So, very fast time to delivery based on all those elements.

Jason: That's awesome, Bruce. So I wanna pivot a little bit, and we're just start with asking Miles. And Bruce, I want you to take a crack at the same question. And give us your opinion on this. To start, you know, Miles, you mentioned six months ago, 12 months ago, the research looked very different as to how many people were adopting this and, and where they were at. What do you see in the next 12 months, right? Or let's forecast a little bit, what do you see in the next 12 months? Do you expect to be surprised? How do you expect to be surprised? And Bruce, same question, right? About how AI and finance and, and the world that you live in.

Miles: I think, you know, right now, if I look at our research and I look at the MIT scissor research, you know, there's probably 30% of companies that are ready. So I think we'll get up to that 30% number who have successfully implemented. There's this other 70%, which are that I like to say, they're either in silos and spaghetti, or they're in bandaids because they bandaided together connectors and all kinds of things. And it's, it's still kind of a mess. Some of those may have some limited success 'cause they have some level of integration, but you know, those who are are behind are gonna start to see some real consequences. So our hope is, is that organizations are gonna figure out, you know, how to get their act together and start industrializing.

I mean, there were a couple of things I really liked that, that Bruce said, and, and one of them was the, his slide where he talked about the different stages. Now that's been around for a long time, but there's a top level which isn't shown there, which is, you know, it used to be that we'd use machine learning decision science to basically come up with, okay, here's a projection, but now we're really able to go not just a projection, here's an action that it can go take. So one of my theories, and I've written an article on this that appeared, I guess about three weeks ago where we talked about, myself and this professor from the University of Porto about the potential to take an thetic AI and give it an optimization model and have it be able to come up with perfect outcomes. I think that's really important.

Another thing I liked, by the way, coming back to Bruce, is I'm, I'm as a Star Trek fan, he basically talked about a universal translator. The notion that I don't have to be an expert and I can go and ask in English what were sales for last quarter, and it can know some of the specifics that I work with my idiosyncrasies. That potential's huge. And so I mean, that, that's what I heard from Bruce just a few minutes ago. That's what they're trying to do. So I think that's gonna be huge potential too. So I think I, my my hope is that we're gonna see some of that 70% number start to go down more to 50%, and then we'll start to see bigger adoption. But I think that's the only thing that can limit this market is people getting their act together. Love it.

Jason: Thanks Miles. Bruce. Thoughts?

Bruce: Yeah, so one thing that's, that's hindering it is all these stores of these companies coming out saying, we're laying off 10,000 people because of AI. So that's not the message to send, that's not how you implement AI. That's, to me, a very poor approach, which is, which is absolutely not helping. We need to, it needs to be a more positive message. We're using AI to enhance our business and enhance our employees. So that's one key message that I really send. It's, it's not, it's, it's not to replace you. So, and that's a big hill to get over, but once you get over that hill, you start explaining these concepts of this is how it works, this is what it does, this is the difference between this and this. And people, then they start like, oh, I get it. And then once, now, once people now start to understand it, and you get past the fear of, hey, this is to, to replace our jobs, then you get people starting to ask, okay, great, I get it. I'm not as nervous now, how can I use it? How can I use it to, to improve operations, to leverage my business, to get deeper insights?

So I think it's a kind of evolution of getting past the fear, understanding what's going on, to be an, an intelligent implementation of it, not we're gonna lay off half our company to implement it. That doesn't make any sense. And then once they understand it, they're gonna ask questions, how can I do it? How can I leverage it? And it's the people with the ideas to say, well, I know how you can do it, so let's take some processes, let's get some quick wins. They start to see it, and once the, you'll start turning and you get that momentum and they see, oh, wait a minute, I see I can do this. And guess what, I'm not spending four hours a day doing these annoying processes that I would like. I'm now spending four hours a day doing much more value x. The people, they will understand that. They get it, like they're smart. They go, I understand I'm doing something a little more interesting. I understand I'm adding more value to the business. I understand I'm helping my career. So, and then once that starts there, you then you get the, oh, but can it do this? Can it do this? Can it do this? And that's when those ideas then start and people get over the fear, they see how it works, they understand how it works, and those ideas start flowing. That's when the momentum really starts and when you really start cranking. So it's a cool thing to see, but it, it's an evolution of the process and, and you, you have to get there before the gates really open if you wanna do it right.

Jason: Love that. Bruce, I know from my background in software design before coming over to product marketing, one of the things, there's a couple of metaphors we'd use. One was, Mario and the fire flower, right? Where like Mario goes from being just like the puny and then he picks up the fire flower and all of a sudden he's chucking fireballs, you know? And we see that as like, software can do that for people, right? And I think AI can do that for people. One other word that I've used before is, bionic, right? A bionic human is like this, you know, a cyborg, right? Or someone who's augmented by technology. And I think AI really has the potential to do that. Not to replace humans, but instead to augment them, turn them into these, you know, bionic humans that can do more interesting and more amazing things than they could before. And that's really what we've tried to do with technology throughout at Domo. But with AI, we have this whole new opportunity, a whole new world, a whole new set of tools and capabilities we can bring, like the universal translator that Miles mentioned. We have all these things we can now do for people.

So Bruce, you hinted at something and I've got two questions. The first one where we can keep these brief is myths. And I hear a lot of myths about AI. I think this idea that like, AI is here to replace humans, right? And, and replace jobs. That's one of those myths that I think that we're, we're maybe busting a little bit today and showing how people can use AI to augment and, and move to more valuable work. What other myths do you, Miles, Bruce, what are you seeing out there that your research, your experiences say, Hey, actually that's, that's not how the world works. It's actually more like this. What would you say to our audience as to the myths and how to bust them?

Miles: Yeah, I think, I think it's a mistake to just think about eliminating employees now. I mean, there, there may be some employees, and Bruce can attest to this as well, who may not wanna come along for the journey and, and that's their decision, but we should provide the training and capabilities so that they can be, you know, savvy on AI and start to be part of the, the journey. Another thing I'm gonna recommend, I I reviewed the book, probably over a year ago, but David Remer came up with a great book called The AI Savvy Leader. And what's so powerful about the book is it's for executives. So if your executive is just thinking about, I'm gonna cut costs, go buy them this book, hand it to them, because David argues for what Bruce is arguing for augmentation. How do I make the job something that is, you know, more valuable? And, and, and those journeys are, are really quite important to, to, to do and, and to go through. And, and how do you make people not scared of, of the fact that they're gonna be having this capability to make their jobs as, as I said, more interesting.

Jason: Awesome. Bruce, thoughts on myths?

Bruce: Yeah, it's, I mean, as, as with many commercials, it's just, and I hear people say, you just do AI, just, you could just AI's just gonna read your data. It's gonna be magic. Well, it's, it's not that you have to have highly structured, reconciled, enriched data. It's, you don't just throw AI at junk and get miracles. So it's, it's definitely not like that. And, and I agree with Miles, it's not here to replace people. AI, like technology is a lever. It's a very powerful lever, and on the big enough lever you can move the world. So, and, and levers are mechanical advantages. It allows you to do more, the things that you wouldn't be able to do. So you think of it as that it is your lever, it is your mechanical, it is your technical advantage of how you can change things faster and better and quicker, the bigger your lever is. And it's a pretty big lever. So, basically that it's just, it's just really, it's not just, it just works. It's, it's really, if you take, and the basic concepts too, you don't need to spend months doing this. You don't have to be a data scientist, but just understanding those basic concepts, why I lay out our strategy and, and some of those basic concepts, understanding that goes a long, long way, to getting understanding it and getting past the fear. But once again, also to, it has to be the, the leadership has to have the right mentality and the right idea. So

Jason: Last question for both of you, comes down to, we have an audience here with us today. They have a lot of different places that they're at on this spectrum of AI adoption and, and where they're at. So it's a hard question, but if you had to choose one thing that you would recommend that everybody should start doing today to move themselves further along this journey, what would it be? What would you point them at and say, try this, talk to this person, try this thing, experiment. What advice would you give on what they should do starting today? Miles,

Miles: I got to interview four CIOs a few weeks ago, and one of the things that came out of that discussion, which I think, you know, it's a little bit of motherhood and apple pie, but it, it really is, is the basics. Start with what are the business problems you're trying to solve? You don't have to fix all your data. Day one, you have to fix the data that's gonna allow you to solve a business problem. So, you know, pick some early wins. Don't pick things that are easy just because they're easy. Pick the things that are gonna be material to the organization. Get that wind in your sail by being successful at fixing that data and, and going and solving that problem. And then you can move on to the broader agenda items. This means that you have to be very coordinated. You have to be connected at the hip to your CEO, in order to be successful and you need more people like Bruce. Several years ago there was a great book that talked about how many organizations have red people who are technology people, and blue people who are domain experts. And what the book suggested is you want to get more purple people. So Bruce is clearly a purple person because he's able to, to talk the language of finance and understand the technology and how it can solves. So Bruce is the kind of person you need to create in.

Bruce: Yeah, that's funny too, because I worked at HomeAway, account, thought I was a developer and and development, thought I was an accountant, and I just said, sure. So, correct it, it is limited in the middle. It's understanding both those process, but that is correct. It is, it's, it's getting an understanding of, where you stand. How does your data look? Is it, is it a mess? Are we close? Do we have the right tools and applications? It's getting an understanding of what exactly each thing does. Like my example is if I want to use data science for generated accruals, accruals, predictive, well that's machine learning. The AI is more the prescriptive to tell me kind of what we do it. So it's understanding where you stand. It's understanding exactly how this works. So you don't say, well, I wanna solve this with that. It's like, well, that doesn't do that, that does that. So just putting all that together, here's where we stand. I understand at a high level of what the different types of data science does. And then do we have the people and technology we think that can get us across that line? And if we don't, well, you need to, you need to find that quickly because a lot of companies are finding it. And as the days go by, you're gonna find you're gonna fall further and further behind. So it's, it's doing that quickly and getting a quick assessment or getting the people in here who can assess it very fast to see where you stand and get that path forward.

Jason: Thank you Bruce. Thank you Miles. I'll give my piece of feedback is if you're trying to figure this out and you haven't yet, come talk to us. We have an amazing data science team. We have an amazing ACE team, just our advanced customer enablement team that just do things like this all the time. And even if it doesn't result in solving the problem the way that you think it is today, just finding different things and getting advice from people, jump on the chat, ask questions, right? Reach out to Bruce on LinkedIn, I'm sure. But there's so many things. We've got Domopalooza coming up here in March where we've got a whole bunch of sessions around AI. Just lots of things to learn about. Go out there, learn something, spend a few moments, get some time set aside to read through something like the Dresner report, to read the book that Miles suggested, right to go and watch an example and see how somebody else is doing it. And I think Bruce, you know, you and I have talked a lot, just experiment, right? Go and build something. Go try to make a workflow. Go try some different prompts in a gen AI tool. Just try different things and see what works. And if you need advice or expertise, you know, find people who have done it before. Just latch onto them, ask them questions and, and try things and experiment.

Well, thank you both. This has been an awesome session. Hopefully people are getting so much out of it and are learning things that they can use to move on this journey and start doing more with AI. Really appreciate both of you being here with us. Can't wait to see what happens next. We'll have to do a reprise here in about 12 months, see if our predictions are accurate, and where we're at. But, thank you. Anything else to say? Anything to share?

Bruce: No. Drop by a Torchy's if you're in the neighborhood.

Jason: That's right. Well, thank you both. We appreciate it and we'll see you all soon.

‍