[Title slide 1. Blue CAI company logo with tagline “We power the possible” appears in middle of screen. Company website www.cai.io appears at the bottom center of the screen]
[Title slide 2. Multi-color background with text centered in the middle of the screen that reads: “Virtual Event: 2024: The year of actionable AI”. The white CAI company logo appears underneath of this text towards the bottom of the screen]
00:00:08 - 00:00:53
Christina Kucek
Hello and welcome to the final session of our CAI Learning Series, the Three Cs of Intelligent Automation. For those who may have missed the previous sessions of our series, you might be wondering why we call it the three Cs of Intelligent Automation. First, the series is brought to you by CAI. We're a global technology services company with a 40 year history of combining our dual strengths of talent and technology to deliver lasting results across the public and commercial sectors. The second two Cs represent your hosts for today. I'm Christina Kucek, Executive Director of Intelligent Automation at CAI. And this is my colleague Chris Zumberge, Executive Director of Technology Services at CAI.
00:00:54 - 00:01:38
Chris Zumberge
So the purpose of this learning series is to take a deep dive on everything intelligent automation such as practical use cases and technology advancements that drive both efficiency and increased productivity. Throughout the series so far, we've discussed many AI topics to educate you on the what of AI, but now it's time to cover the how. So sit back, grab a drink, and prepare to learn how to apply AI in real world situations such as thinking ahead to your 2024 strategy. If any of you guys have questions, we encourage you to ask them in the chat. We will try to get to all your questions as they come up during the session. But if at any time you want to learn more, please visit our website at cai.io for articles, client success stories, or to set up a discussion with someone on our team.
00:01:39 - 00:02:12
Christina
Welcome everyone to today's Automation Learning session event 2024, the Year of Actionable AI. My name is Christina Kucek. Briefly about me. I'm passionate about assisting clients and their automation journey from building hyper automation teams for RPA and document extraction to machine learning and artificial intelligence solutions. Our solutions drive efficiency, cost savings, and a competitive advantage. With me is my co-host Chris Zumberge, who's better known around CAI by as nickname Z.
00:02:13 - 00:02:45
Chris
Hi, everyone. I am the Executive Director of Technology services at CAI and I have been designing and delivering AI solutions for over five years at CAI and in addition to helping execute on CAI's overall AI strategy, one of my areas of focus has been the ethical use of these publicly available machine learning solutions. So I am very glad to be with you today as we discuss this AI evolution and how we will transition from 2023 that the year of conversational AI to the year of 2024 actionable AI.
00:02:46 - 00:03:02
Christina
All right, let's get started. As Z mentioned in today's 30 minute discussion, we're going to talk about how 2024 is the year of actionable AI with our special return guest Bhavin Shah. So let's go ahead and introduce him.
00:03:03 - 00:03:26
Chris
Bhavin Shah is the CEO and founder of Move Works Enterprise copilot platform. He is an entrepreneur with over 25 years of experience taking companies from inception to scale. In 2016, Bhavin co-founded Move Works, the generative conversational AI platform that unifies all enterprise systems. Bhavin began his career in educational toys at Leapfrog Enterprises, so welcome back.
00:03:27 - 00:03:30
Bhavin Shah
Thank you Z and Christina, thank you for the warm introduction.
00:03:31 - 00:04:20
Christina
Yeah, of course. Thank you so much for being here with us today. We only have 30 minutes, so let's jump right in. In 2023, we saw the rise of conversational AI in the mainstream. We've been all over the country, the three of us, explaining technologies like generative AI, large language models, neural networks to all kinds of interested people and educators and business executives, county commissioners, all kinds of stakeholders. Looking to 2024, Z and I were tossing around predictions and the one that seemed most compelling is what he referred to is actionable AI, which I immediately loved. So my first question is to Z. I'd love to hear you define what you meant by actionable AI and how do you see that as a key driver in 2024?
00:04:21 - 00:05:15
Chris
Yeah, absolutely. When I refer to actionable AI, I'm referring to the transformative role that AI can play as a personal assistant in your daily work. So an AI system that actively assists you in completing your work. And a lot of people will hear that and say, well, we're already using enterprise generative AI solution to extract insights from data to help me compose the first draft of an email, review my code or thousands of other that we're hearing about. But when I talk about AI helping you complete your work, I'm talking beyond providing insights or answering questions. It's when AI becomes your assistant in getting work done. So that means that you need to teach the AI to do the work that you do. It needs to be able to perform that work inside of all the different systems that you do work in, whether that's your HCM system, your CRM, your ERP or any of the other tools that you use to do your job.
00:05:16 - 00:05:59
Chris
It needs to understand what you mean when you say, clean up all my old opportunities in my CRM, right? And then it needs to know how to connect to your CRM to make those changes for you, whether it's using an API connection or sequencing instead of RPA scripts to make that happen. And for 2024, I think that we're just at the perfect launching point for this discussion because a lot of business users who are not technologists, who aren't part of it are more interested in AI and automation than they've ever done before. And thanks to the popularity of ChatGPT, whether you use it or not, you know about it and it's taken the concept of AI from being a buzzword or a blockbuster movie villain is something that they can actually use.
00:06:00 - 00:06:22
Chris
And it got people thinking about generative models and asking the question, how can I use this to solve my business problems? And I think that as people dream big about the answer to that question and then learn about the limitations of generative only chat platforms, they're going to be very interested and very receptive to adding capabilities to their chat platform.
00:06:23 - 00:06:24
Christina
Awesome, thanks.
00:06:25 - 00:06:33
Chris
Yeah, I would love to hear from Move Works themselves. I mean, do you agree that [inaudible 00:06:31] one of the breakthrough successes of 2024?
00:06:34 - 00:07:42
Bhavin
Yeah, I think you really articulated it well in terms of where things are headed. And if you think about the world that we've been in over the last 12 months, let's take March as a point in time. Everyone was talking about RAG, retrieval, augmented generation, and essentially the narrative and the concepts that started to explode were how can you use large language models to better derive insights from data? As we all know, large language models are trained, they're fixed in time based on what they have. And the way you overcome that is by bringing in additional data sources to the form of a vector database and you vectorize these documents and you give them to the large language model in the form of a prompt giving it additional choices, additional information to consider. And while that is important, and that will continue to be a key feature that will go through its maturation over time, I think the point you made, which is actions are really where you start to drive a lot of the business value and the impact to business efficiency.
00:07:43 - 00:08:44
Bhavin
So I think that if you're an AI system that just does RAG, you're going to be woefully challenged over the next coming years to stay relevant. And I think that sort of is the framework here, which is businesses are now investing in generative AI. I've seen it ourselves. There's actually budget starting Jan one for this category that didn't otherwise exist for a lot of organizations. And starting Jan one, there's also a expectation that these investments will lead to certain business outcomes, business efficiencies, changes in how teams operate and such. And so for that, there's higher levels of scrutiny as to what is this solution going to do? How is it going to reduce our costs and how that all plays into this. And so I think the action piece of this is where a lot of the value will start to emerge. And I think for us, the enterprise employee service space, right, is a very complex environment.
00:08:45 - 00:09:14
Bhavin
We have to connect to hundreds of different systems, systems of record so that we can answer employee's questions to start, but then we can also take action. And I think that's sort of where we're seeing a lot of the needs arise. And I think 2024 will really be that. And so we definitely agree. I think there's a whole breakdown of this problem that I can kind of go to a little bit more, but does that make sense?
00:09:15
Christina
Absolutely.
00:09:16 - 00:09:47
Bhavin
Yeah. So if you think about what generative AI is doing that we hadn't had the ability to do before, I'll take the problem in two halves. You have the problem solution and then, sorry, it's a problem space and this solution space. So if you go back and we were one of the first companies to put the [inaudible 00:09:37] came out in 2019 and it did a pretty good job of extracting entities and finding out the intents for a particular piece of text.
00:09:48 - 00:10:48
Bhavin
And so we had a pretty good handle on the problem, what the user's asking for, what they need help with, but the solutioning was always a bit rigid and it would be us connecting to a certain API, but that was a fixed workflow or that was a fixed set of capabilities that it could do. What that sort of left us with was a very good understanding, but a limited number of things that the system could dynamically involve itself with. And so when you think of generative AI, I think what it does is it gives us the ability to create more dynamic actions, obviously knowledge generation by bringing in different pieces of information and then combining that and then of course follow-ups and clarifying questions. And so we actually come up with an architecture that's part of this thesis that everything is going to turn into actions and actions are going to take a bigger and bigger role is that you need a system, especially one like ours that spans across the entire enterprise, one that's extensible.
00:10:49 - 00:11:46
Bhavin
So our new use case, we need to build that or we need to add that to the smarts of the system. You also need something that has steerability. Why? Because as much as large language models are creative and they hallucinate, you also have to control for that. In certain cases, you want the model to always do X. And so having steerability forcing it down certain paths is key. Interoperability is also I think one of the most interesting parts of this, which is in the world of traditional actions and recipes and Ansible tower scripts and this and that, you kind of had to pre-design every single type of action that was taken. Whereas now with large language models, we can combine them in ways that were not possible. I'll give you an example. If I go to our bot today and say, "Can you give all my direct reports access to Figma?"
00:11:47 - 00:12:34
Bhavin
Okay, that's a brand new query our system hasn't been programmed to understand. But with large language models, problems, solution, we break that down. We understand who I am, who my reports are. By taking action, we have a whole plugin architecture that allows us to go look at Workday, look at the various HRIS systems to say who reports to [inaudible 00:12:09], then identify who they are in active directory or G Suite, then go and check to see do they have access to Figma already? And those that don't, then it can tap to those APIs. If you think about this, this was never really possible because you sort of have to come up with that query ahead of time. And then pre-action stuff, I didn't even touch on this. In a real enterprise, you need to look at permissions, you need to look at ACL rules, business policies.
00:12:35 - 00:13:29
Bhavin
And what we've been fortunate having sort of been doing this for a while, is we're able to train these large language models to take these actions with greater levels of performance because we're tracking. When Bhavin asked this question, how well did we do? Did we resolve the issue? Did we close out the issue? That data then comes back, leads to better outcomes, and factoring in the permissions, it leads to this flywheel of efficacy that starts to increase. I think you're a hundred percent right. The world of actions is where you're going to see a lot of investment dollars. You're going to see a lot of companies try and solve this, and it's a much more complicated problem than what you might be seeing out there in the consumer world of, hey, I can add a plugin to ChatGPT and that'll solve everything. I think unfortunately, the enterprise world is just way more complex and requires these other elements to properly detect.
00:13:30 - 00:14:19
Chris
Yeah, I think you touched on one of the most amazing things that people are thinking about and talking about, which is kind of that when we think about automation and the traditional sense of it, writing a script, RPA is very prescriptive, right? You take a process, you take it from end to end and it kind of has this... The large language model, this new way of thinking about that, of saying, all right, if I can ask it a question, I can ask it to do something that's never done before. If it understands my ecosystem of RPA scripts, can it take them, separate them into micro automations and reassemble them into what I'm asking to be done? So you can enable your workers to run automations that they've never asked for before, right, without it having to go back to an RPA team, maybe to re-engineer from beginning to end on top of being able to say, "Hey, I need you to do this one-off thing and do it at 4:00 AM tomorrow."
00:14:20 - 00:14:55
Chris
And being able to schedule it conversationally through there. And for some of these one-off processes that you would maybe like an AI would assemble in real time, you might not get the ROI on that to have a whole team put it together. Maybe it's something you do for maybe an hour or two once or twice a quarter, right? And so that cost benefit analysis to putting that in place maybe wasn't quite there. But with AI in place now you can just run those things and it enables, I guess, more [inaudible 00:14:50] right? Or the citizen automation engineer being able to make that happen.
00:14:56 - 00:15:28
Bhavin
I think it's a classic phenomenon that we've seen in technology. The more you can lower the price and the cost of doing something, it doesn't actually reduce the number of people required per se. It actually increases the appetite of consumption. So here you described some cases where you wouldn't have had the ability to automate them, so you would just do it, you'd live with that low grade pain and you'd proceed. But now a lot more can get automated and so you're going to end up looking for even more use cases and more opportunities. So yeah, a hundred percent agree.
00:15:29 - 00:16:12
Christina
Yeah. And then you also mentioned rules, right? Like implementing rules and make sure that you're following business rules and policy rules. I think that is going to be one of the things that actually hinders deployment of generative AI is trying to jump through all the hoops to get people just comfortable with the idea of allowing actionable AI to take action, like granting access, especially to systems like HR systems, protected data, finance systems, and kind of baby stepping in to those waters. And like you mentioned, forcing them down a specific path I think will help build trust and put the guardrails on to protect people from themselves, some of their messy systems that maybe don't have enough controls or access controls in place.
00:16:13 - 00:16:53
Bhavin
Well, it's a harder problem to solve than most people think. And you've seen some new announcements about enterprise bots having challenges with that in the last week, some big announcements and leaking permissions, leaking documents that... And so I think there's a lot of work when you talk about verifiability having fact-checking models, going back through the loop again to verify these things are correct. These are layers and layers of different systems that have to be built up over time so that you can have that confidence that it's always going to follow the rules.
00:16:54 - 00:17:47
Chris
Yeah. And it begins to kind of beg the question of when we're putting these models out here and we are getting towards actionable ones, so let's say one that can send emails on behalf of me or anybody else, being able to have the ability to say, [inaudible 00:17:08] explainability. This is how you came to a conclusion. This is the data you pulled about maybe a client or a partner or something. You assembled this email so that I can... You said the fact checking models on the model side, but me as a fact checker on the output because if it comes out of my account, I'm responsible for it. So I need to have that kind of the transparency into how it's being made, the data being pulled. And you talked about retrieval augmented generation techniques, and they're coming out with a lot of more interesting add-ons that with chain of notation, which is almost like how do you build up the citations that prove out the Wikipedia sources of the degeneration that was created here?
00:17:48 - 00:18:02
Chris
And I think that companies and platforms that are bringing that to the users and making that transparent and clear and explainable, they're going to find a lot of success, especially when, as Christina was saying, going over that hump of trusting AI.
00:18:03 - 00:19:05
Bhavin
Yeah. And look, technically these models are opaque. There is no real yet research to show how to explain how these models all perform the different reasoning that they do. But the way you overcome that, the way that we've managed to kind of navigate this is by breaking down these tasks into sub tasks across different models or different steps that then give us steerability effectively, right, so that if we know the output of this task led to this other task and that's where there was sort of a miss, then we can go and inspect that and explain, okay, this is what ultimately happened. So I think that the idea that you just give this to one big model and it just solves it all may have some merit in some future state, but right now it doesn't give you the controllability, the steerability that is required.
00:19:06 - 00:19:35
Christina
Yeah. One of the things that you guys both touched on I think is the ROI factor in explaining... The tech is so cool. We get that, but we need to find a reason to pay the bills, right, to implement the tech, not just for technology's sake, but to really move the needle. So Bhavin, I think that actionable AI has a really clear and obvious several use cases where the ROI is very easily seen.
00:19:36
Bhavin
Yeah.
00:19:37 - 00:19:44
Christina
What are some game changers that Move Works is planning to roll out in 2024 that you would like to showcase during this time?
00:19:45 - 00:20:43
Bhavin
Well, I a hundred percent agree. I think that we are very focused on delivering value to our customers, and that's sort of really aligned with their goals that they have. Every business is trying to transform itself. I think what we see is sort of three main pillars, right, drive operational efficiency, which hopefully is pretty self-evident, enhanced experience. Everyone wants the workplace to be a nicer place to work, a more seamless place to work, an easier place to get things done. And then the overall AI transformation, evolving the business systems, evolving the structure of how data is stored so that AI transformation can actually have a bigger and more enduring impact within the business. And so some of the innovations that we've invested a lot in and we continue to is one is our knowledge studio. So if you think about any copilot, it's only as good as the data that you are able to provide it, right, and data matters.
00:20:44 - 00:21:37
Bhavin
But what if some of the answers that people are asking questions about don't exist right now? But they may exist in work notes, they may exist in unstructured repositories of information. And so what our system does, the Knowledge Studio, is allows you to analyze all of that unstructured data, IT ticket notes, resolution notes, et cetera, and then brings it all together to create new content that you don't currently have assembled. It also does something even more interesting, which I think is... It also will look at the articles you have, scan everything else that you have, and find inconsistencies and gaps because there may be seven different articles about your travel policy and which kind of hotels you can stay at and not, and being able to say, "Hey, look, there's multiple answers out there. Let's pick or verify one document over the others," is key.
00:21:38 - 00:22:33
Bhavin
Of course, at the core of all this is our next gen copilot, one that we've been talking a lot about, but it really is a much more conversational interface than anything you can find for the enterprise today. And it can handle multiple requests. It can sort of keep track of all of these actions plus information that it needs to assemble. So think about ChatGPT, but tailored to your enterprise, ensuring that there's no hallucinations. Everything is cited at... To your point Z in terms of that we don't quite do the chaining of things, but we go back to a source that we have been given knowledge of is verified by the internal team, and you can validate the factuality of it. So that's at the core. And then there's of course what we call Creator Studio, which is allowing our customers, and I talked about this earlier, the extensibility, it's key, right?
00:22:34 - 00:23:28
Bhavin
We offer a lot of out of the box capabilities. Why? Because the world is oftentimes very homogenous. You'd be surprised to see the same kind of IT issues, HR issues coming out of a hospital as we do a video game company, as we do an industrial company, as we do a media company, right? And so that sort of starts off our customers with a ability to see that business impact minute one. But then historically, it's been very difficult to extend these systems to do new things. And where that becomes even more complicated is they start to collide. Looking up a ticket inside of your ITSM may look like the same kind of question of looking up inventory inside of SAP or Ariba or one of these other systems. And so what customers of ours, like Toyota North America, Micron, Peloton Networks and Honeywell and others have done is they're building these use cases on top.
00:23:29 - 00:24:17
Bhavin
But what's unique here is we're able to use new techniques of few shot learning and generative AI such that they can create 15 new use cases, put them into production in a couple of days, versus what they had historically seen with some of these other toolkits that would take a few months to see that come to fruition. So I think you're going to see everything accelerate. And I think that all of these drive back to that thesis, Christina, right, of what's the business value? What are the use cases, what's the impact? And 2024, I can't wait. It can't come soon enough because we're going to see a huge adoption of all of this technology that has been kind of floating out there and getting announced, but now it's really going to hit home.
00:24:18 - 00:24:46
Christina
Yes, we've been using Creator Studio actually to extend our use cases because we found out what people were asking through the tool and said, all right, let's connect to different data sources. Let's create some new flows, for lack of a better word, some new solutions for them. So that is pretty awesome. Z, I'm going to ask the same question of you because we're deploying a lot of AI solutions at CAI. What are you most excited about rolling out?
00:24:47 - 00:25:40
Chris
I mean, for me, we actually touched on it a little bit and that's why I was so excited to talk. Actually, I think changing the way that we look at the way we automate processes, going from these kind of in the software world, like the monolithic solution, right, into almost like that microservices approach for automation so that we can start to look at having AI sequence, how to complete tasks, whether it's an API call, a piece of automation here, a piece of an automation here. And that's really what I'm the most excited about. I mean, I know I can speak for myself, I won't speak for you guys, but I have a lot of small admin tasks, right, almost low value ad work that I need to get through. And I think that most everybody has that, right? And I think we're also at a time where everybody is being pushed to be more efficient, and we have these tools and we're being asked to optimize the time that we spend.
00:25:41 - 00:26:26
Chris
And I think that being able to then create these AIs that can sequence this kind of small automation for everybody, almost like we all have our own little intern or own little assistant with us at all times, is really going to just change the way we interact. I mean, at work. And I'm personally absolutely the most excited for that. I can't wait to see how people in finance are like, I asked it to do X, y, Z thing. It was able to figure out all these stuff and put together and check it out like a little show and tell of what everyone was doing around the company. And I think that's really just my 2024 New Year's wish.
00:26:27
Christina
Awesome.
00:26:28 - 00:26:31
Chris
Yeah, I think-
00:26:32 - 00:27:32
Christina
I was going to chime in about what I was excited about. Well, you guys just had me thinking, and some of my clients have disclosed when we talk about AI solutions, they're like, "We just don't have the data." We have the data literally all over the globe, and we haven't made the major investment to get all of our data into a data lake. We don't have the BI layer really available. And these are global companies that do a lot of business. So really the ability to use generative AI and point it to multiple separate data sources to allow them to just ask questions about their data. And I have just this vision of seeing executives that I work with regularly sitting in some sort of meeting and somebody asks a question, they say, "Well, hold on a second." And just typing it in regular English expression about what was our sales, how many widgets did we build last year, or how many people are deployed in this area?
00:27:33 - 00:28:12
Christina
And just being able to answer that question. Right? I mean, I think we've all lived through the major BI reporting, we're building cubes for business stakeholders, and this data's important. I mean, it's awesome to do that, but then when they don't use those reports and they're not looking at it and they're not, it's like, well, this is close to what I wanted, but I kind of wanted it different. You go through all these iterations, I'd love to start at some of these clients with let's give you access to the data and find out what kind of questions you're asking, and then look at how we're going to do that data manipulation and consolidation from that lens instead of the other way around.
00:28:13 - 00:29:11
Bhavin
Yeah, and you bring up, I mean, it's an evolving space, but if you think about access to information, what's the right architecture, right? Everything's moving towards this [inaudible 00:28:26] architecture that we're seeing in the enterprise. That's something that we've architected our system around. And for some use cases, you're going to want to bring that data into a data lake and you're going to want to then query that. But the truth is most systems are quite federated, and most enterprises are going to continue to be very fragmented. You may have a dream that everything comes together in one central place, but that almost never happens. And it's a very tall order because things tend towards entropy. So even if you get it all in there today, four months from now, there's now data showing up in other places and you're constantly trying to chase that. So I think what people are doing, and we're one of those is instead of fighting that, we have these plugins that can talk to these various systems.
00:29:12 - 00:30:03
Bhavin
And then the real sort of engineering innovation that has to occur is automatic plugin selection, knowing which system to go to, when to go to it, how to disambiguate, and then sort of do what's called pre-triggering work to figure out do you have the right permissions? Is this going to lead to an answer that does give us something that we can work with? These are all kind of complexities. If you think about any of the early evolution of copilots you probably have tried, most of them are like, okay, connect up to three plugins and then you can ask me questions about those systems. But the challenge is when I say, "Hey, can you add me to the sales deal?" First of all, you have to know who me is. You can't send me to an API. API call, function call me as an argument, and then sales.
00:30:04 - 00:30:57
Bhavin
That's too generic. And so the systems actually have to go figure out who you are, who then what sales means in this context. Oh, it's a distribution list that goes to Office 365, and I think that's where people are getting stuck because it is a hard problem. But if you can build the right sort of [inaudible 00:30:26] framework where these different agents are all working independently, but also in conjunction with one another, you get this quite amazing infinite explosion of possibilities to your point of, hey, you're in a meeting and you have a question. Well, that question may be a lookup in that data lake, but it also may need to get some other data out of another system of record as an API call to convert something before inspecting that data lake, because not everything is normalized.
00:30:58 - 00:31:28
Bhavin
And knowing to do that, even in some recursive fashion, in some cases, we watch our system do that, and it's pretty exciting when we see it on its own recursively saying, "Oh, okay, now I need to ask this other question." And it comes back and pretty explicit about it. So it shows you the work that it's doing. I'm reading these eight articles. Okay, now I'm summarizing the articles. Now I'm going to come back here. Now I'm looking up this information. And I think that's where the future is taking all of us.
00:31:29 - 00:31:37
Christina
I think it just happened. We just connected the dots between the ROI, that value proposition and the really cool technology.
00:31:38 - 00:31:56
Chris
Yeah. I would love to dive down the rabbit hole of multi-agent simulations with you as you're starting to touch on that there. But we are definitely running out of time here. So I would just want to say it's been an absolute pleasure chatting with you, all of you, but especially Bhavin. And I would also like to thank our audience for your attention and your participation.
00:31:57 - 00:32:33
Christina
And we hope you'll take away these three key lessons from today. Number one, AI will expose your dirty laundry. So make sure you do your diligence in protecting your data within your organization. Two, actionable AI will require deep understanding of your business processes, and you're going to need to establish trust with your stakeholders. And three, in a global marketplace, you have an option to move away from ETL jobs running all day, all night long and move to more real time intelligence.
00:32:34 - 00:33:07
Chris
And later, we will be sending this to everybody who has attended a recording of this entire chat, so you can share it with your colleagues or your peers. And in the meantime, if we didn't answer your questions organically here and you're interested in learning more about CAI Intelligent Automation, or you know someone who is, please visit our website again cai.io and fill out a contact form. Feel free to contact any of today's speakers via LinkedIn. Our direct LinkedIn pages were shared earlier in the chat. And thank you. I hope you have a wonderful rest of your day.
00:33:08 - 00:33:19
Christina
Thanks to everyone for joining. We hope you have a great holiday season ahead. Stay tuned for more events in 2024 as we continue to ride this AI wave together. Thanks, Z. Thanks, Bhavin.
00:33:20 - 00:33:29
Chris
Thank you.
[Closing slide 1. Blue CAI company logo with tagline “We power the possible” appears in middle of screen. Company website www.cai.io appears at the bottom center of the screen]
[{"asset_name":"www.cai.io","asset_url":"https:\/\/www.cai.io","asset_type_code":"site"},{"asset_name":"Resources","asset_url":"https:\/\/www.cai.io\/resources","asset_type_code":"page_redirect"},{"asset_name":"Events and Webinars","asset_url":"https:\/\/www.cai.io\/resources\/events-and-webinars","asset_type_code":"page_standard"},{"asset_name":"2024: The year of actionable AI","asset_url":"https:\/\/www.cai.io\/resources\/events-and-webinars\/2024-the-year-of-actionable-ai","asset_type_code":"calendar_event_single"}]