 Welcome everyone. My name is Richie Bavasso, and I'm your host today for this exciting event. I'm also the co-founder and CEO of NQ Medical, a machine learning computational biomarker that uses passive data collection from your personal device to manage neurological health. Thank you for tuning in to this just tech connect and utilizing artificial intelligence to advance your startup. Today's program is broadcasting live as an official global entrepreneurship week event. You can follow the rest of this week's activities on social media by searching hashtag GEW20202020. Artificial intelligence, or as we all refer to in its AI, is the ability of a digital technology to perform tasks commonly associated with intelligent beings, such as rational ability to reason, discover meaning, learn from previous experiences, et cetera. Large companies have been using AI to great benefit for quite some time, but smaller companies and startups are also able to use this technology to build and scale their innovations, in some cases, to a more efficient degree than the larger companies. AI also includes such terms as machine learning, deep learning, and big data. Today's web chat will be a discussion from leading experts in their respective fields on artificial intelligence, who will share their perspectives and thoughts on the future of AI. It's important, and we really want to engage you and take your questions during this program. So please prepare yourself, ask your questions in the chat space next to the video player, and or use Twitter, hashtag just tech connect, G-I-S-T, tech connect, and we'll field as many as possible. Let me introduce our panel of experts. First starting with Frazier Kitchell. Frazier is the co-founder and CEO of KEF Robotics, where he leads the corporate, the company's corporate strategy and business development. Frazier, could you give us a little bit about your background in your company? Sure, good morning, thanks Richie, and thanks everybody for joining. So KEF develops aerial autonomy for aircraft. So what we're trying to do is develop software that can help fly an aircraft without a human pilot. And what we focus on is not building the aircraft, but just providing the autonomy. So you can think of us as a driverless car, software company for aerial purposes. So we leverage AI and machine learning in a number of ways. We're trying to, we're in a computing limited space, and we're trying to teach an aircraft how to perceive the world like more like a human would perceive the world. So we're leveraging machine learning to help us understand what objects look like from the sky or perform a task like recognize a building from above. So we're a small team, we found it about two years ago and I'm excited to tell you about our experience with ML and AI software to date. Thank you Frazier, autonomous cars are certainly an exciting area where AI is being applied. Our next panelist is Lakisha Hall, Director of Transformation and Strategy for IBM. Lakisha, welcome and let's hear about your background in your work at IBM. Sure, so thanks so much Richie and thank you all for inviting me today. As Richie mentioned, I lead transformation and strategy for our FSS unit, which is our financial services sector. So working on our marquee clients in the financial markets, banking and insurance industries. But particularly around AI since 2016, I led our delivery organization or professional services that implemented data and AI solutions across the world for IBM and its clients. Very unique to our team is that we not only implemented solutions for AI, but we also could teach the client how to do it. We most famously said, we could do it for you, we could do it with you, or we could just teach you how to do it. Great, thank you. So you'll see we get the small company in the big company perspective during this webinar. Next up, Vishal Kapoor, who's a principal in Deloitte Strategy Analytics Practice. Welcome Vishal and once again, a brief overview of your background and the work you're doing at Deloitte. Thank you, Richie. Good morning, everybody. I am a principal in our strategy and analytics practice and part of global AI and data management practice that Deloitte has across the globe where we serve Fortune 500 companies as well as large government and public services. So clients in the use and adoption of AI, I've been sort of, my experience is a balance of commercial work that I've done for half of my career in the last 10 to 12 years. I've supported federal government and state and local government agencies in the US in a wide range of AI solutions and adoptions. All right, thank you, Vishal. We're gonna go to questions to our panelists to get started. I was once told that webinars are like a buffet. Everyone wants to get up and eat, but no one wants to go first. So to get us going and we're looking forward to your questions in the chat space, we're gonna start with Akisha. Akisha, what are some of the benefits of implementing AI that you've seen in the market, both intentional and unintentional? Sure, sure. So we like to talk about an AI journey and I'll talk a little bit about that later, but what we've seen as intentional benefits, our customer wanted to be more available, right, to their clients and particularly important in this environment, when we can't have that face-to-face interaction as much as we want to, that availability became ever more important, right? We wanted to be, our clients wanted to have more reaction to their clients as well, to increase customer service, making sure that their brand delivered the right answers at the right time every time. We will also want it to reduce wait time and make their employees more productive. Lastly, our clients really wanted to make sure that they were able to leverage their clients and their customers and their employees in the best way possible. So kind of removing some of those mundane tasks and enable them to actually work on more complex works in their business. But the items that were unintentional were the kind of no collar, as Jenny Rometti called it, new collar jobs that came out of this. So it was increased, right? Employee satisfaction out of these new jobs that developed when the cognitive era started. I remember a quote from one of my clients and she said, you allowed me to become more inquisitive and we love that because they were able to not focus on these mundane tasks and strolling through mounds and mounds of data and then start to work on more complex items. Another thing that was really unintentional was the reduction in nutrition. So I was working with a large company working on their call center. And they noticed that not only were their employees getting up to speed more quickly, but they were staying longer because they had more opportunity to do different things in the organization. So all of those things were tremendous, right? Value to our clients. Thank you. Yes, some would say we're still in the paleontothic era of artificial intelligence and there's far, far more to go. Vishal, what are some of the bigger commercialization trends and benefits of being an early adopter and how important is it to be first to market? Yeah, so I think it's very interesting actually delighted a survey, you know, we do an annual survey of AI adoption across, you know, 4,500 sort of clients, large organizations across the world. And last year when we did that, you know, AI was a sort of core imperative in terms of first to market and as a differentiator. But that opportunity window is closing in as more and more organizations start adopting it and it becomes almost like a de facto, you know, as a way of doing business. It's starting, you know, it's sort of becoming, you know, less cutting edge in many contexts and becoming more mainstream, which would reflect for the adoption of AI. You know, in terms of commercialization opportunities, I think there are, you know, if you just, you started with the definition, right? Which is it a lot, you know, broadly speaking, it is allowing me to do things which humans thought were only human possibilities, but now machines are able to do, whether that's, you know, I find, you know, sort of detect patterns or I visually see I interpret language, I'm able to, you know, see patterns in video and identify objects, whether machines can do a lot of that, then inherently it can have a game changing impact in how we do work. So it essentially transforms the way we do work. And that is through industry after industry and government agency after government agency. You look at any industry and any government agency, you will see, and if you see it from the lens of what AI can do, it is a different future. So we have to sort of imagine that. So commercialization opportunities are immense, whether, you know, startup companies are looking at domain-specific applications of AI to affect in those domains, whether it's banking or life sciences or, you know, sort of energy in government space. There's a lot of fintech and red tech kind of innovation happening. So there's domain-specific commercialization opportunities and then there's capability-specific commercialization opportunities. So, you know, I can create, you know, computer vision-related capabilities that are sort of not domain-specific, but apply them to the domains. I can create natural language of individual solutions. I think we'll see significant increase in AI ethics and AI quality-related, you know, capabilities being infused by vendors. So I think commercialization opportunities are immense because it, you know, it's becoming more than a hobby. It's becoming, in fact, it's actually becoming more than a differentiator than the first move with advancing, becoming mainstream. And every domain that we can think of is going to be significantly impacted. So I think the opportunities are immense. Right, thank you. Just a quick follow-up question. Is there a vertical that's farther ahead? Is fintech farther ahead than medtech? Or is there one specific area a vertical that you think is doing well? I think, you know, it's hard for me to be very specific, but certainly in banking and finance, you know, the applications of AI are tremendous. You know, we've seen banking and finance use, you know, advanced analytics and statistical modeling for years. And it is just but natural for them to sort of get into the cutting edge of AI usage. Pharmaceuticals are, you know, looking at, you know, significant application of AI. You look at millions of dollars being spent in, you know, clinical trials and, you know, that entire sort of ecosystem, you know, a process. So significant gains there. So I think life sciences, healthcare is a big, you know, focus, banking would be another focus. And I also say that some of the toughest challenges facing that have to be tackled out by government agencies. You look at the volume of data and the kinds and the significant challenges government agencies have. They have a significant, you know, imperative and, you know, we are seeing a lot of traction in that space as well. Right, thank you, thank you. Top three, you know, among all the industry sectors. Thank you, thank you for that. Question for Frasier. Startups face many challenges that have just inherent in being a startup, but are there particularly issues related to startups in the AI space? And if so, are there specific tools that can be used to improve on these efficiencies and work past these challenges? Yeah, so I'm excited to talk to the audience today, partly because I know there's a lot of entrepreneurs out there, so we are a brand new business. We're actually not even two years old yet. And we founded without, we are a R&D team, but we didn't have any direct experience in ML or AI. We didn't have to use it right away. So I did wanna relay at least three points to small companies. The first is that there are a lot of free assets out there for training. So we are developing algorithms in-house to, like I said in my kind of introduction, teach an aircraft how to perceive its surroundings or recognize a power line for being a power line or a person for being a person. Those are very difficult things for a computer to do, obviously, but there's a lot of free data available. Unfortunately for us, there actually aren't a ton of data sets from the air. So we have a lot of data collection that we take on ourselves, but we're hoping that teams like Microsoft and Google are actually making some of that information available to us, well, to us and to the public. So that's number one point. There are assets available for you. You don't have to gather all the data yourself. And we're talking about robotics now. There's a little bit of a different field than, say, what IBM's probably, at least, Laquisha's been talking about today. The second point I'd make is that we, I think, we're still at a point where you can learn on the job to some degree. So if you can attract talent, there's a lot of younger people coming out of school where AI and ML are very attractive fields and they're studying and there's a lot of innovation going on, so it's a very dynamic space to be in. So we have had some success with in-house development of talent in our two-year history. So that maybe speaks to where we are technically, where we're a little bit more on the research side than on the implementation side, but as we move into implementation, we think our employees can catch up and basically be getting close to the state of the art in their fields. So I would encourage entrepreneurs to not be so intimidated by AI and ML that you think you can't do it yourself. This, in the end of the day, is still a field that people are inventing as we go forward. There was a question about streamline and efficiencies that we've realized and I think this is true of robotics in general, but also of AI and ML in general, which is we're offloading human tasks to machines. So one thing in the drone space, when you're flying to an aircraft, especially the small aircraft that we fly, we're talking sometimes four or five pound vehicles, you're extremely limited in power and computing. So you are wasting time and power by sending data back to say a human operator to analyze and then say proceed to the next action you should take. So we're using ML to basically streamline our efficiency of a mission or of a shipment of a flight because if we can do the processing right at the point of the observation. So if our cameras are taking in an image of a human and we know that aircraft knows that it has to make a safety diver right away, that's a time saver and that's a mission enabler. So like has been said, we are offloading the burden on a human and what that allows us to do potentially is sort of, we'll have more flights, have safer flights and have one human operator, maybe overseeing 10 aircraft at a time as opposed to one pilot, one vehicle, which is the current model. Great, super. We're getting questions from online. So thank you very much for the audience for sending us questions. And I'm gonna jump back to LaKeisha for a moment with a question. Search marketing, target display, social media ads are three of the most common ways to incorporate AI. Other others. Oh, absolutely. I mentioned before that, we really want to take the clients on a journey and what we've seen in the journey is that our clients really start off with kind of these FAQs, these question and answers, making sure that, you know, their clients understand what question and answers they can provide initially. But then the clients move to transactions. What can we automate? What is simple to automate first and then go on to more complex tasks? They then move to expert assist is what we call it. So training their employees to make decisions faster, to find information more quickly. And then as they get more advanced, they go into offering products and services via AI. They also start to look at what we're seeing now is going into regulatory compliance, helping to make decisions, right, on compliance issues. So we've seen a complete evolution, right, from things that are more, I think, customer centric to really transforming their businesses overall. Perfect, thank you very much. Question for Vashal from the audience. What does AU not do well and what's your companies not look to use AI for? Oh, so I think AI inherently is predictive in nature, right? So it has this notion of probabilistic prediction. So depending on the problem, you have to calibrate what is the right appropriate use of AI in your context. So applying AI to detect whether someone has cancer or not and then recommend the right sort of sort of treatment regimen is one thing. And then deciding the risk on a particular claim that has been filed against an insurance company are two very different sort of problems which require different levels of accuracy and precision. So understanding, so that becomes important. I think if you have a very high degree of explainability that is needed, some of the inherent deep learning capabilities that exist today will cause you some challenges. So be mindful of that. So I would say I think while the use cases are widespread, there is also a lot of hype. I would say for businesses and organizations, I think they need to root the decisions of where AI should be applied, not just for AI's sake, but really what is the business impact that you're actually going to create, right? What is the office's efficiency you are going to create? I think the conversation needs to start there. AI needs to enable that, but the conversation needs to then again end there which is did I actually create that impact a lot without that orientation? There are a lot of false R and Ds and failed efforts which can happen and there's a risk of that happening because there's so much hype and media tension and talk about it. But I think kind of back to basics, what is this impact that I'm going to achieve and then appropriately use the right AI technology to provide that. There are certain solutions where AI doesn't make sense and you should not be just jumping to that conclusion. But having that careful strategy and evaluation is a core part of charting out that path through a speak of, you know, we're in the top of it. Very good. Yeah, Rich, I'll actually jump on that if you'd like to. So yeah, we're in the aerospace world and aerospace is a traditionally very conservative industry and it should be. It should be, you know, the FAA has, and the US has almost a perfect record of aircraft safety over the last decade even. So while we are excited about what AI and ML are offering us, we are in the research phase with that. And we know, I mean, it's going to be very difficult to get AI or ML powered flight into the national airspace. So we use it as a tool to extend what are kind of more understood, understandable methods in the field of navigation. So we would take in, you know, we couldn't just fly with cameras right now. We would need other types of sensors to get into the airspace. So some of the words that Vishal used were great. I mean, it's probabilistic. Like it's uncertain. You don't always know what your output's going to be. And there's tons of examples where AI is fooled by very, very basic things that even a younger person would have no trouble understanding as a hazard. So yeah, in robotics, while it's promising, a lot of people are still fairly frustrated by it and there's still a ton of work that needs to be done. So it is not here yet. It's hyped and it's really interesting. It's really powerful, but there's still a lot of work to do. Great. I'm going to editorialize this next question a little bit because as Frazier mentioned, we have a lot of entrepreneurs in the audience. I myself am an entrepreneur. My company has pursuing four different revenue strategies because revenue is a lifeblood of any startup organization and we haven't committed to any of the four. So my question, this question from the audience is too Frazier, but the person who asked is also very interested in Laquisha and Vishal's response. Let's talk about how do I devise a revenue strategy with AI based on your experience? And there's also seems to be a move for the big tech companies rather than growing their AI initiatives internally, they're making acquisitions. So how might both the revenue strategy and the acquisition from a larger entity affect to how I or any entrepreneur in the audience go about executing on their business? So Frazier first, because it was presented to you, but Laquisha and Vishal, I'll be interested in your comment. Sure. Okay. So I know we have a very international audience today and I hope I don't give a two U.S. United States centric answer here, but to date, so in our two years, we have found most of our funding through governments, the government agencies. So our two primary funders are their department of defense and NASA. And NASA's obviously interested in aerospace and space. So we have some work there. And the department of defense is highly interested in AI and ML and they're helping us develop a lot of our capabilities. That doesn't mean we're talking about purposes for war. We're actually doing more like a warehouse inspection type things. And I put that caveat at the beginning of my answer because I hear oftentimes from people in Europe or people even in Asia that their governments aren't as maybe proactive in kind of seeding small businesses. So we're able to access a lot of small business innovative research grant money that isn't always available overseas. So I'll say this, it's really nice to get government contracts because you don't have to basically borrow money from investors or bootstrap yourself, bootstrap meaning start up with your own capital as you're starting your business. So far we've found that strategy to be effective in building from just a three person team to an eight person team. And we're not sure how long we can continue on just those contracts. We'll have to come to market at some point but it's given us time to innovate while we get there. LaKeisha, do you have a comment relative to revenue generation and the acquisition targets from large companies? Yeah, so what we've seen is companies really leveraging savings from putting in different initiatives and actually repurposing the workforce to actually go in and invest in their innovations. That was incredible feedback that we've gotten from a number of partners. From the position though of startups and acquisitions, it's really an opportunity. It's an opportunity I think someone spoke to scale before was an opportunity for some of the smaller companies to get to a scale that they had not reached before. I think in having discussions with some of our industry leaders, they're seeing this change in the marketplace where maybe 25 years ago, a pharmaceutical company has identified these smaller companies that they've leveraged that innovation, right? We're seeing that now that we're leveraging that innovation through AI and they're not building it themselves. They're leveraging these smaller companies to do that. So it's really a win-win in that area. In addition to that though, there's a need to reach the customer at the point that they're making the decision. And I think AI is doing that more incredibly than we've seen before. And this is really a game changer both from a small company to a large one that's actually using that innovation. Thank you. Gashal, any commentary? Yeah, I think I'll add, so I mentioned if I sort of am a startup in this today's day and age and I'm looking to kind of create a solution and use AI, I will really look at two options, right? If I am going at it from a domain specificity standpoint, I'm addressing a particular market vertical and a particular domain specific solution using AI, that's one way, but having that sort of laser focus on what that target market is. And in other ways, I'm basically building AI capability which can broadly be applied into different scenarios and different sort of verticals. Two broad classes of solutions out there. And when you look at sort of hyperscalers, the cloud providers and how rapidly they are evolving their capability, at Google and Amazon or Azure, they are evolving their AI capabilities sort of native capabilities in the cloud. I think the domain specific of startups are hard to replicate because it hadn't required that domain understanding and knowledge which takes years to build. But the capability specific, that's a real concern in terms of the acquisition sort of what you mentioned, that's a real concern because that capability is evolving by these large hyperscalers I call them or we call them at a much rapid pace. But again, thinking through the revenue models, AI, if there are certain companies who are finding the niche in that space, utilizing and knowing how to work with the various cloud players and adding a value added capability in conjunction and compliment to that and offering that as different deployment models, sort of license models, as a service models, hybrid models, I think there are all these different models that AI, because many clients are gonna want to just consume it by the drink, some will put in a commitment and buy sort of in an enterprise fashion journey. So I think supporting those different revenue models would be key for startups to have that flexibility in how they make revenue. But I think having a laser focused on the domain or the problem plus, I think that would be key to identify. Right, thank you, thank you. If you don't mind, I'm gonna add a comment as well as a in the trenches, roll up your sleeves, CEO of a company that operates not only in the US, in the EU, in the UK and in Asia, what I'm about to share I think is universal. Number one, no one cares about your technology. Okay, they wanna know what problem you sell. So I think if you present your company and you're offering your platform as a solution provider, I think you have a better opportunity, at least a more opportunity for an engaging conversation. And my second word of wisdom is, don't expect the customer to get it on their own. You almost have to treat them as a kindergartner and explain to them why your offering is valuable to them. What problems that sell for them specifically, what pain point? I can tell you, I've done it the other way and they had lots of great conversations that went nowhere. And when I began to drill down on, what is the problem that the person in front of me is facing and how can my AI solution solve it? It's a much more lucrative result. So I'm sharing that from the voice of wisdom, pain and experience. Okay, quickly, I'll just take a break for one second. I'll get back to the questions. For those of you who are just joining us, you're watching the GIST Tech Connect on Global Healthcare Innovations in partnership with Deloitte. Our experts here to answer your questions, so please keep generating more questions in the chat space and we'll get to them as soon as we can. All right, going back to the question sheet, this one is more focused on employees. So within your organization, have employees or potential hires raised issue with how you, your organization, excuse me, is using AI? Are there any ethical issues that have been brought forth by your staff and how do you deal with those? I'll start with Akisha and then go to Michelle because you represent larger organizations. Yeah, so I think we're a bit unique. So we haven't had any ethical issues raised as far as AI is concerned from our employees, but we certainly are on the forefront of making sure we're leveraging AI for good. And as you probably have noticed, we have pulled out of the facial recognition space as a company, so that's really our focus. We are leading the industry in AI bias understanding, our models and being able to help our customers understand that, and I think this is why we're able to get this adoption from our critical banking and financial market clients because we need that transparency across the board. So I think it's really two-fold, one, making sure that we are able to explain what our AI is doing and then making sure that we're leveraging that for good overall. Excellent. Michelle? Yeah, so a couple of very recent examples come to mind in the last six months alone, right? So we are supporting a few stated local agencies using AI to predict and prescribe which households, which individuals in the jurisdictions of those states are at high risk of COVID in terms of hospitalization. So if they get affected, they are most likely to go into hospitalizations and have significant impact. So being able to have states sort of show up their supply chains and hospital beds in those locations where you have these vulnerable populations. Now, the classic conversation there was very, very clear interpretation of the results. So you can have a degree of extendibility behind those decisions because now you're basically making decisions with millions of dollars on where to sort of push your resources, where to set up test sites, for example. So explainability was key to be able to generate a level of confidence with the stakeholders. Another one, which is more about bias in quality of algorithms is something we're supporting states in unemployment insurance frauds. So there are billions of dollars being paid on claims over the last six months. And we all know there's significant fraud happening in that sort of ecosystem as well. So using AIML to identify fraudulent claims and then be able to put a high risk score on them. But then you also need to do that in a fashion which let's just say, in the probabilistic nature you don't wanna get into a scenario where some legitimate claims are actually part of those false positives. So again, a high degree of focus on EI quality because what is at stake is people sort of unemployment benefit. So I think those two scenarios are sort of live in my mind clear because we encountered them in the last few months where the appropriate use of AI and the right focus on explainability and quality was key to actually using it in the field for real decision making. Great. Sticking with the, thank you, Vishal. Sticking with the theme of human resources. Question for Frazier. What's your experience in securing talent in this space? And the machine learning, data scientists, et cetera. AI talent, how do you get support from someone with experience? How do you find people to support your product? Sure, finding talent in a small business is always really difficult. It takes a ton of my time, much more time than I would hope for. But so we are, I'll just say we're looking at people who have a little bit on the less experienced side, I would say. So we were in a pretty innovative field where we feel like we need people coming out of university, maybe masters, maybe PhD programs who are, who have just freshly versed in what's the state of the art. So it does take a long time. I'll also say that we were talking actually before this call went live about working from home and what that's changed. We have a hard time because we're in Pittsburgh, Pennsylvania and the US, which is a smaller town. It's got some robotics heritage, but it isn't one of the big cities like New York, Boston, West Coast where the talent's really heading. So we do have to get people excited about what we're doing, which is trying to, first of all, bring aerial autonomy to more people and more aircraft makers. But also, can we leverage AI ML and these exciting ways that will push them to have a really exciting career and motivate them all while they're working? Yeah, I mean, there's lots of follow-up questions I could ask, but I wanna stick with what the audience is asking right now. And I'm shifting gears a little bit. And I think this is a big small company worry, all right? And I'm gonna leave it up to the three of you to decide who wants the answer or you can all answer it. How do I minimize the risk of someone stealing my idea or copying me, either a larger organization or an organization more aggressive than I am as a small company in the global markets? It's kind of a intellectual property question. Yeah, I'll go first, because we think about this often as I'm sure the others do as well. But I hope this isn't an intimidating answer, but our basic idea right now is that we can't protect our IP. We don't go for patents, partly because we're a software company, we don't really believe we can protect those things. But also, if that's your idea as an entrepreneur starting out, you might need to change your mindset a little bit. You're gonna have to constantly get better and people will steal your idea and that would be a compliment to you. You need to build a team and a culture that's innovative enough to stay ahead of the competition. And I mean, that is the tax you pay for being in business and publicizing what you do. So yeah, IP concerns are real. You probably won't be able to defend yourself from anything as a really small company. Your time's gonna be taken up with all sorts of other things and just go forward. That's what I'd say. Yeah, I think first the market is your best bet and staying innovative. From my perspective, I share with you, Frazier. Lakisha, Michelle, any additional wisdom in that regard? Yeah, so from our perspective, we certainly do go after our parent patent is quite aggressively and we certainly have a huge right focus on IP protection and certainly our clients' protection of IP as well. So I would say it's definitely a top concern but I wouldn't let, from a small business perspective, I wouldn't let that deter you, right? Go out, continue to innovate. Like Frazier said, be first to market, right? And then find partners that you can work with and continue to innovate. Great. I agree wholeheartedly with what Lakisha and Frazier said. I would also add, I think in today's day and age, the notion of IP is very different than what it was 10 years ago. So contributing to open source frameworks, having some part of your capability, in fact, be out in the open so there are more users and more developers, actually more technical staff using it is an interesting part of the strategy that you can play while preserving some part of the IP but some of it making it a bit more openly accessible so that there's widespread usage because that itself can lead for you to be sort of even more valuable in the ecosystem. So I think it's, but I think the core of it is if you think you have an IP today, well, someone somewhere in the next few months is going to have something similar. So yes, it is incumbent on you to continue to innovate, continue to advance and invest in that. Yeah, keep in mind, I'm sorry, go ahead, Michelle. Having a strong team, motivating them, keeping them energized, I think this is part of the game so that they continue to advance as a whole history. Yeah, you as a small company have on your side momentum and history. And when we start getting worried about the competition, we tell ourselves, nine women can't make a baby in a month. Anybody that wants to copy you still has to go through the gyrations and the market pain of figuring out, how do I build this product? How do I sell it? And you're way ahead in that regard. So I agree with all the comments that were shared. Stay out in front, stay innovative and you'll do well. I wanna ask you a question now because either this is one person or two people are asking about farmers. That's farmer with an F and any innovations in the area of farming and anything that would help farmers generate more sales. And I don't know where to start folks, I'm assuming it's not you Frazier, but feel free to answer. I'm happy to give it a thought. But this is not my, I am a outside the office. I love farming. I kind of wish I was a farmer a lot of the time. But yeah, so yeah, we're in the drone space, agriculture is huge. There are plenty of applications for aerial autonomy in drone farming. And I'm not trying to pitch our tech, that's not where we actually work. But so there's a lot of interesting things going on with data collection. So if you can, if you can fly drones through an orchard, and you can look at the individual plants, maybe these are trees, you can figure out how to feed them or whether they're showing signs of pest infestation. So that's all ML powered, right? A human can do that quite well, but a machine can automate most of those steps and do a very good job at it. So I don't know who's farming what on this call, but if you're some of those high value ad products, fruits, things that need to be on the supermarket shelf, absolutely, there's tons of things there. And then actually I'll just jump back and get off on a little bit of a tangent. But as a former biologist, there's all sorts of interesting things going on in biology for creating new crops and figuring out how to adapt crops to climate change and all sorts of stuff. So yeah, there's data powered science going on everywhere. And I would just add a little bit, so everybody wants sort of can we produce organic produce and the right use of insecticides, there's this notion of precision agriculture. So if I can, based on the climatic conditions and the particular crop you're growing, the soil conditions, variety of factors, I mean, I'm just, you know, just creating some hypothesis that based on that, you could, if you could sort of determine how much insecticide to use precisely rather than sort of one side that typically happens, there's a lot of avenue there. I would think supply chain, you know, how can you preserve, you know, certain goods for longer again, using a variety of those factors, I think those could be impacted by AI as well. So I think I know for a fact, USDA is starting to look at some of those capabilities as, you know, as they launch sort of new programs to help sort of the farming community and application of AI can have the impact in that space. Yeah, absolutely. And that's exactly what I was gonna say as well. Vishal, you know, the food safety, right, is a tremendous space, right? So understanding where your food came from, protecting it from the farm to the table. Great. If it's okay with the organizers that you can text me, we started five minutes late, I'm gonna plan on ending five minutes after the hour, unless you tell me otherwise. A lot of great questions coming in, one that I think is, you know, top of mind for all of us involved in this space and it's an open question to all the participants, data-driven security, right? Security in our data, how can startups ensure or any company ensure that the data they're collecting and using for their AI-powered tools is maintained and kept secure? Professor Stotter-Lakeshia in IBM is probably very concerned about that particular area. Yeah, absolutely. So we are very specific about not using our client's data to enhance our models, to improve our models. So this is of utmost importance to us. And it could be a range of topics on keeping your data secure, either where it is, whether it's on-premise, whether it's in the cloud, et cetera, the location of your data, meaning physical location or country. So we keep all of those standards in place as well. And then also making sure that your whole company understands, right? How you're managing, controlling, accessing the data. And we certainly have a number of offerings now, right? With our cloud packs, et cetera, to be able to manage your data appropriately and making sure that it's in the right place, you know where it is, and we can manage the operations overall. So what I would add to that is, so in my experience, the two biggest roadblocks in making significant adoption of AI is one is the cultural barrier. So people are used to doing things in sort of old ways and they just have to get out of their heads to imagine the new future. But I think the other one is this, you know, concerns about security, you know, organizations, especially when it comes to organizations collaborating on certain data assets to create some new solutions, the inherent challenge and the roadblock they have is, how is this data going to be secured? Who's going to get access to it? How are they going to use it? What's the purpose they're going to use it for? Is this going to be appropriately utilized? So and so forth. Security, if we, you know, we need to tackle that, you know, that challenge to be able to really scale AI. And specifically to AI, what it means is more maturity in solutions that de-identify some of the data sets, meaning eliminate the PII or identifiers, but still allow me to use that data for learning and then feed it back into systems where, you know, I have those identifiers. So being able to scramble data sets, being able to mask data sets and use that as part of my sort of AI ML pipelines, I think it's going to be a significant requirement as we continue to achieve it. Yeah. And as we review, right, our data architectures, you know, we are constantly looking at what data do we actually need, right? You, you know, there's a wealth of data out there, but what do we actually need to solve the problem? And you may find, right, that this entire data set is not what you need to be able to solve your problem. And then where, like Vishal said, where you need it, you can appropriately mask it. All right, another question has to do specifically with the energy environmental, clean energy environmental space. Could anyone on the panel please share what you see as some most promising advancements in that particular area? Vishal, Vakisha. Hello. Some examples I can share with you in the clean energy space. So a lot of energy companies, as well as the Department of Interior as a sort of regulator, are looking at the use of AI for, you know, a variety of scenarios that were difficult to do before. So for example, looking at satellite imagery to be able to identify oil spills, let's say across, you know, the fine reason across across the globe would be some scenarios. Certainly, you know, you know, Vakisha can comment on hot on climate related, you know, weather related data that they have and you know, sort of climate related models that both NOAA and you know, sort of other parts of the government can use. I think so there's certainly a lot of applications of that. I mean, a little less directly involved in those, but you know, I would love to hear about the possible analysis. Yeah, we have, so things that I've seen and I was thinking about different examples. So the one around just predicting right oil spills, one of our largest clients was Woodside. And so the outstanding work that they are doing in predictive and understanding the knowledge, right? Of how to manage what's happening on these oil rigs at the edge was significant. One of the conferences that I went to a couple of years ago around reducing CO emissions is something that the teams are looking at in leveraging AI. And then also with the supply chain an interesting component that the teams are looking at is how they're tracking the leveraging the use of hydrogen but also tracking where their actual oil gas are coming from in the supply chain and making sure that there is no interruptions there. So that's just some of the examples that I've seen in the industry. Right, unfortunately, I'm not gonna be able to take any more questions. We're running out of time. We wanna give our panelists each an opportunity to share their final thoughts, pearls of wisdom. I will share with the three of you that a lot of these questions are focused on use of tools for communicating with customers, whether in the marketing space, as chatbots. So if appropriate in your final thoughts, if you perhaps can weave in a little information relative to that subject matter, I think the several people who are viewing today would be appreciate it. LaKeisha, let's start with you. Sure, so the conversational AI or chatbots are actually my sweet spot. So there is a master class that I created with partners within our team that we can certainly send out and let you review. But what's very key about that is just knowing your data, making sure that you have great data for your teams to use, making sure you have your SMEs on hand and continuously learning. I would say would be the key things for a small business. And I think Frazier spoke to that. And it's even applicable to a larger enterprise because we saw that we prove adopt and scale. So if we start off small and we prove that something successful, then we can gain that adoption. So with our larger corporations, it's with the different lines of business, but with a small company, it could be with a partner. And then you're able to scale that out to other lines of businesses or to other partners if you are a small company. So I would say prove adopt scale. Excellent. Frazier, I'm gonna go to you next and I'm empathetic because we're both small business owners and speaking to the entrepreneurs in the audience, what are your final thoughts? Well, I'd echo what I said in the beginning and maybe tried to hammer on throughout because you have to dive in the tools and the amount of data and the things that you can leverage as an entrepreneur in a small business are getting better all the time. And they continue and Vishal noted this as well. They continue to be more open source. And so things are more complex than they've ever been before but they're also more attainable than ever before. So I would just recommend diving in, finding partners, finding mentors. You don't be afraid to ask anybody for help in this world when you're an entrepreneur more likely to not provide it. Great, and never give up. Sure. Yeah, of course. That is key. That's right. Vishal, your closing comments. Yes. So I would say I think we started with sort of AI having a material impact on the way work is done in the future. And that is true across industries as they look to increase the top line, better products and services and more access to consumers or bottom line squeezing out more efficient use of resources and cost and dollars. Figuring out your niche as to which problem are you going to solve? What industry vertical are you going to address? Or which problem class in AI are you going to enable with a degree of sophistication and capability that is complimentary and lacking today? Understanding the effects and bias and quality of AI are going to be paramount in core to successful solution in the future because I think the hype is translating into real results and real results will only happen when we are able to solve the ethics and bias and quality control issues. I think thinking of ecosystem plays, meaning think of yourself as a player in the ecosystem that you operate in and how you can work within that ecosystem I think would be a good mindset to adopt for new startups as well. But I think we haven't even seen, we are at the very early stages of this maturity and this transformation. So I think that's good news for a number of the entrepreneurs who are looking at possible revenue targets and possible ways to grow their companies. I think we're at the very early stages and a lot to do on the next couple of days. Frank, thank you. Yes. Well, thank you, Lakisha, Frazier, for joining us and sharing your expertise and your experience. I would have any one of you or all of you on my advisory board. So great feedback for the audience. Thank you. I also want to thank everyone who joined us today, especially the viewers around the globe. I hope you got something from this exercise and let's keep the conversation going. Even though today's program is done, we don't want to stop the conversation about entrepreneurship. If you want to continue the conversation on Twitter, you can go to hashtag GIST Tech Connect, that's G-I-S-T Tech Connect. Check back here for more information and some follow up at gistnetworking.org. I believe this will be posted shortly for you to access for information on upcoming events from GIST and other programs like this. Once again, on behalf of the panel, thank you very much. And enjoy your day and your week and keep on entrepreneurial. Thank you. Thank you. Thank you.