View Recording: Top OpenAI Production Use Cases for Revenue and Operations

Nathan Lasnoski Welcome everyone. 0:2:51.960 –> 0:3:3.430 Nathan Lasnoski Welcome for this next hour we are going to have an awesome time talking about the top production use cases for Azure AI and other AI use cases within your business. 0:3:3.710 –> 0:3:23.710 Nathan Lasnoski I’m going to show you how companies are getting started with this, how they’re making progress, how they’re moving into production, and I’m going to walk through in production use cases showing you the example, walking through how they did it, give you some ideas to go back to your business with that are going to help you accelerate your movement into getting value from AI. 0:3:24.200 –> 0:3:27.550 Nathan Lasnoski So just to introduce myself a little bit, I’m Nathan Linowski. 0:3:27.560 –> 0:3:29.610 Nathan Lasnoski I’m concurrencies chief technology officer. 0:3:29.800 –> 0:3:32.700 Nathan Lasnoski Been working with AI for about 8 years. 0:3:33.150 –> 0:3:51.50 Nathan Lasnoski That eight years has been an awesome journey, helping customers see ROI from this, and it’s been accelerated in a completely insane way around the last year as companies have really seen the light bulb of what AI can do for them and have been excited about turning it on and making it a thing within their organization. 0:3:51.490 –> 0:3:53.300 Nathan Lasnoski And I’m looking forward to talking more about it. 0:3:53.310 –> 0:3:54.30 Nathan Lasnoski So let’s dive in. 0:3:55.760 –> 0:4:20.240 Nathan Lasnoski So what you can see in front of you is a representation of how AI has changed and will change how the market has changed over time in respect to the GD production of GDP and what’s really interesting about this is how mass production, transportation and communication has changed the way that we were able to produce goods within the world. 0:4:21.390 –> 0:4:32.950 Nathan Lasnoski This opportunity that AI presents to us, I would say, is underestimated by most business users underestimated by most business executives. 0:4:33.90 –> 0:4:34.980 Nathan Lasnoski Now you can say like wow, how is that possible? 0:4:34.990 –> 0:4:38.60 Nathan Lasnoski Like, how could you possibly be underestimating AI at this point? 0:4:38.110 –> 0:4:39.340 Nathan Lasnoski Like there’s so much hype. 0:4:39.350 –> 0:4:40.560 Nathan Lasnoski There’s so much buzz. 0:4:40.570 –> 0:4:42.970 Nathan Lasnoski How could the hype not be big enough? 0:4:43.230 –> 0:4:53.780 Nathan Lasnoski Well, in a sense like the height might be big enough, but when I talk to typical executives, they’re thinking about value streams or thinking about opportunities to enable their business with artificial intelligence. 0:4:53.890 –> 0:5:2.20 Nathan Lasnoski They’re thinking about the mission of their business, but sometimes they’re not thinking about every individual within their organization is going to be impacted by this. 0:5:2.670 –> 0:5:12.530 Nathan Lasnoski Every individual in our economy is going to see change as a result of artificial intelligence, in the same way that smartphones change the way we work, the Internet changed the way we work. 0:5:12.540 –> 0:5:27.180 Nathan Lasnoski PC has changed the way we work, mash production and changed the way we work and this is our opportunity, not only to look at the future of our businesses through the lens of AI, but also the opportunity to look at the future of our people through the lens of AI. 0:5:27.510 –> 0:5:36.980 Nathan Lasnoski So as you leave this conversation today, think about the mission of your business and think about how the mission of your business will change as a result of enabling artificial intelligence. 0:5:36.990 –> 0:5:49.30 Nathan Lasnoski As part of that picture, but even more so, think about every person that enables the mission of that business and how they will be enabled by AI and our responsibility to help them make that journey real. 0:5:49.680 –> 0:6:8.320 Nathan Lasnoski And as you see, as we go through many of these use cases, you’re gonna see how people’s job roles are changing as a result, not being eliminated, but changing as a result, being forced multiplied as a result, enabling them to gain their ability to use their creativity to achieve good within their mission. 0:6:9.440 –> 0:6:12.370 Nathan Lasnoski So I want you to to know three things. 0:6:12.700 –> 0:6:16.750 Nathan Lasnoski The first thing I want you to know is that the AI opportunity is for real. 0:6:17.200 –> 0:6:18.790 Nathan Lasnoski This isn’t a science experiment. 0:6:19.20 –> 0:6:23.830 Nathan Lasnoski It’s not a situation where we’re just talking about doing things with AI. 0:6:24.120 –> 0:6:32.560 Nathan Lasnoski This is an AI opportunity that we’ll have real and lasting impact on the operational savings and revenue production of your business and the context of its mission. 0:6:33.660 –> 0:6:45.950 Nathan Lasnoski What you’ll see with the examples that I’ll show you are companies that are getting real value and those are opportunities that you should bring back into your organization to be able to translate those ideas into real value for your business. 0:6:46.50 –> 0:6:46.850 Nathan Lasnoski And we’ll help you do that. 0:6:48.590 –> 0:6:53.240 Nathan Lasnoski Which you can see here as well is that operational savings tend to be the low hanging fruit. 0:6:53.650 –> 0:6:57.280 Nathan Lasnoski The reason why they tend to be the low hanging fruit is because they’re there. 0:6:57.450 –> 0:6:59.80 Nathan Lasnoski They’re there to take advantage of. 0:6:59.120 –> 0:7:8.940 Nathan Lasnoski They’re there to drive efficiencies from they’re processes that you’re running right now that if you really had to do it again, you would change the way that you executed on that process. 0:7:9.170 –> 0:7:13.30 Nathan Lasnoski And a lot of times the data is ready to be able to gain those operational savings. 0:7:14.310 –> 0:7:26.600 Nathan Lasnoski The third thing is that revenue production is a very real and tangible opportunity within most organizations, both incrementally doing the same thing you do right now faster and also within the disruptive space. 0:7:27.130 –> 0:7:31.450 Nathan Lasnoski So take these tangible things to know and bring them back to your organization. 0:7:32.620 –> 0:7:36.870 Nathan Lasnoski We buy all these scenarios that will walk through as a component of this presentation. 0:7:38.700 –> 0:7:46.700 Nathan Lasnoski So I want everyone to take just take a second and putting the chat or just kind of contemplate this to yourselves. 0:7:46.710 –> 0:7:47.450 Nathan Lasnoski You put in the chat to see. 0:7:47.460 –> 0:7:52.650 Nathan Lasnoski Even better is where are you as a business in this adoption cycle? 0:7:53.590 –> 0:8:3.710 Nathan Lasnoski Are you at executive alignment where you’re just getting your organization understanding what the mission of the business looks through like through the context of AI? 0:8:3.720 –> 0:8:5.870 Nathan Lasnoski What are the guardrails we’re placing around it? 0:8:5.960 –> 0:8:11.220 Nathan Lasnoski How do I engage and incremental and disruptive innovation within my organization as a result of AI? 0:8:12.530 –> 0:8:17.50 Nathan Lasnoski Or have you already started brainstorming as a result of that executive alignment? 0:8:17.140 –> 0:8:23.40 Nathan Lasnoski You’re I wanna point where you’re understanding the use cases and how it translates into activity that you’re gonna be performing. 0:8:25.90 –> 0:8:25.580 Amy Cousland Excuse me. 0:8:24.830 –> 0:8:25.820 Nathan Lasnoski Or are you in a position? 0:8:25.620 –> 0:8:26.260 Amy Cousland Excuse me. 0:8:26.270 –> 0:8:26.870 Amy Cousland Excuse me, Nate. 0:8:26.880 –> 0:8:27.20 Amy Cousland What? 0:8:27.30 –> 0:8:32.370 Amy Cousland You know your slides are not are not moving along, so if you can change it to a presentation view. 0:8:33.550 –> 0:8:36.630 Nathan Lasnoski Ohh, thank you for telling me that umm one moment. 0:8:36.650 –> 0:8:37.480 Amy Cousland Sorry. 0:8:37.620 –> 0:8:38.70 Amy Cousland Let’s see here. 0:8:39.760 –> 0:8:42.270 Nathan Lasnoski You know, I’ve had that problem now a couple times. 0:8:42.280 –> 0:8:45.660 Nathan Lasnoski That’s why I was checking that OK, you should see this now. 0:8:46.400 –> 0:8:47.110 Amy Cousland Perfect. 0:8:47.380 –> 0:8:47.850 Amy Cousland Thank you. 0:8:45.670 –> 0:8:49.230 Nathan Lasnoski This is gonna make a lot more sense now that you are seeing the screen. 0:8:49.740 –> 0:8:50.150 Nathan Lasnoski OK. 0:8:50.160 –> 0:8:52.440 Nathan Lasnoski Thank you for correcting me on that. 0:8:52.750 –> 0:8:53.210 Nathan Lasnoski Uh. 0:8:53.370 –> 0:8:54.500 Nathan Lasnoski On on that, Amy. 0:8:54.770 –> 0:8:58.620 Nathan Lasnoski OK so here are back this up just one moment. 0:8:58.690 –> 0:9:0.780 Nathan Lasnoski Here is the AI adoption cycle. 0:9:0.850 –> 0:9:12.600 Nathan Lasnoski These are the three that we’ve just walked through is executive alignment group envisioning and now we’re moving into this idea of understanding mission driven versus commodity use cases. 0:9:12.830 –> 0:9:21.260 Nathan Lasnoski So this idea that there are use cases we’ll talk about today, they’re very focused on direct revenue production and direct operational savings. 0:9:21.450 –> 0:9:30.810 Nathan Lasnoski And in order to achieve those, you have to be precise in the way that you’re answering a question or you’re providing an answer or you’re delivering on prescriptive results. 0:9:31.320 –> 0:9:42.510 Nathan Lasnoski There’s another lane here, which is the commodity uplift and if you want to think about the commodity uplift lane in with an example, that would be something like spell check. 0:9:42.600 –> 0:9:54.890 Nathan Lasnoski OK, so when I write an email and I’m writing that email and I have a word that I’m not sure how to spell, I don’t go pull out my dictionary and find that word and then type it and then continue on in my life, right? 0:9:55.0 –> 0:10:1.400 Nathan Lasnoski I have this auto change that happens to the word that I’m writing as I go, and sometimes it’s accurate, sometimes it’s not. 0:10:1.590 –> 0:10:21.560 Nathan Lasnoski Based upon autocorrecting all the funny jokes on autocorrect, but ultimately it’s a drastic time savings and me writing emails commodity uplift is going to be that times 100 in terms of our ability to produce one of our colleagues like to say never produce a draft again, and this is really the M365 copilot space. 0:10:21.570 –> 0:10:53.310 Nathan Lasnoski It’s other copilots being result released as parts of your SAS platforms that you use in the market and your organization needs to be ready not only to intake those commodity use cases, but then the think about what’s really unique about you and that’s these mission driven scenarios that you pull forward into piloting it, validating it delivers value moving into production iteration and then going after a scaled pattern which is how do I do this again and how do these all fit together. 0:10:53.320 –> 0:11:1.210 Nathan Lasnoski So how does an individual production use case I’ve built fit into perhaps another individual production use case in the same organization? 0:11:1.690 –> 0:11:7.600 Nathan Lasnoski Maybe even solving some of the same problems, but might not have have to have a different kind of answer structure. 0:11:8.110 –> 0:11:11.120 Nathan Lasnoski So the anti adoption cycle fits across this entire. 0:11:11.790 –> 0:11:17.620 Nathan Lasnoski This entire ecosystem, but what I’ll tell you is a lot of companies try to start here. 0:11:19.80 –> 0:11:34.590 Nathan Lasnoski They tried to start with enabling the technology rather than focusing on gaining adoption and alignment at the executive level and by widening the tent in the business to understand the right scenarios to go after within their organization. 0:11:35.900 –> 0:11:39.930 Nathan Lasnoski So as we get started with this, I want you to consider this example. 0:11:40.140 –> 0:11:50.260 Nathan Lasnoski So this is a an customer service page on Choice Hotel group and what you’ll notice about it is that it is. 0:11:50.430 –> 0:11:54.80 Nathan Lasnoski It’s an experience that at one point was pretty transformative. 0:11:54.90 –> 0:12:13.410 Nathan Lasnoski Like we were very used to calling into sort of IVR systems to be when we had a customer service issue and we’d wait on hold and we’d talked to someone and then we waited on another hold and we talked to someone and they built this platform that you can go to their customer service page and put in your question in your name and they’ll call you back, OK? 0:12:13.470 –> 0:12:19.430 Nathan Lasnoski That the callback scenario that was really cool at one point no longer is this state of the art. 0:12:19.500 –> 0:12:24.860 Nathan Lasnoski In fact, this really misses my expectation when I have a customer service issue. 0:12:24.870 –> 0:12:52.640 Nathan Lasnoski What I’m expecting in this scenario is to be able to open up a chat application and be able to start talking immediately in asynchronous way with a human or with a bot that’s going to solve my issue and had a similar example of this where I had booked a hotel room that hotel room was booked for the wrong day opened up that application without having to stop what I was doing and corrected that that that reservation that I had without ever having to get on the phone with a person. 0:12:53.810 –> 0:13:6.820 Nathan Lasnoski So what this really represents to me is you need to meet the customer where they’re at, and you need to meet the customer with where their expectations are in terms of what normative experience should look like or better. 0:13:7.130 –> 0:13:12.620 Nathan Lasnoski And in this case, what they have right here used to be state of the art no longer INS. 0:13:13.740 –> 0:13:15.140 Nathan Lasnoski Now look at this example. 0:13:15.150 –> 0:13:16.770 Nathan Lasnoski This is the Hyundai palisade. 0:13:16.780 –> 0:13:19.370 Nathan Lasnoski Hyundai Palisade is tried to have a chat application. 0:13:19.380 –> 0:13:20.510 Nathan Lasnoski Let’s see what they’re doing. 0:13:20.640 –> 0:13:27.360 Nathan Lasnoski What kind of gas should I use in my palisade and they try to answer the question but they can’t. 0:13:27.370 –> 0:13:29.40 Nathan Lasnoski Can you please say this in a different way? 0:13:29.50 –> 0:13:33.490 Nathan Lasnoski I answer best when asked short general questions and what I say when I look at that. 0:13:33.500 –> 0:13:36.960 Nathan Lasnoski I’m like, aren’t you the app for the Hyundai policy? 0:13:36.970 –> 0:13:41.660 Nathan Lasnoski Like I think chat GPT knows more about your van or truck than you do. 0:13:41.970 –> 0:13:43.540 Nathan Lasnoski Why can’t you answer this question? 0:13:43.870 –> 0:13:52.280 Nathan Lasnoski So if you attempt to go down the road of creating something, remember you’re going to be measured against the state of the art. 0:13:52.570 –> 0:13:54.250 Nathan Lasnoski So we were building an application. 0:13:54.350 –> 0:13:55.580 Nathan Lasnoski This was many years ago. 0:13:55.680 –> 0:14:2.780 Nathan Lasnoski We’re building this application with a search bar and the search bar could only search for things that you spelled correctly. 0:14:3.700 –> 0:14:9.950 Nathan Lasnoski So like if you spelled the word wrong, it wouldn’t find the right destination answer, and they’re like, well, this is working really well. 0:14:9.960 –> 0:14:10.390 Nathan Lasnoski Check it out. 0:14:10.400 –> 0:14:16.260 Nathan Lasnoski It’s like, no, it’s not working well because you’re being measured against who you’re being measured against. 0:14:16.270 –> 0:14:21.730 Nathan Lasnoski Google in the search space and if you’re measurement doesn’t measure up, you’re not going to be looked at favorably. 0:14:22.810 –> 0:14:25.900 Nathan Lasnoski Same thing with AI in the chat space. 0:14:25.910 –> 0:14:34.800 Nathan Lasnoski If you’re not comparable against something that someone’s gonna get from chat, GPT, you’re not going to be looked at favorably in the virtual assistant space. 0:14:35.90 –> 0:14:52.140 Nathan Lasnoski So think about the quality level that’s necessary to achieve success, but there’s one really fantastic thing the bar to get in to have that level of quality has been lowered because the capabilities are becoming more commoditized and that’s some of the things I’ll show you today. 0:14:53.470 –> 0:14:58.480 Nathan Lasnoski So as you think about your ideas, I want you to think about them in two different major buckets. 0:14:58.710 –> 0:15:1.90 Nathan Lasnoski The first bucket is the incremental bucket. 0:15:1.100 –> 0:15:2.860 Nathan Lasnoski These are things you already do. 0:15:2.950 –> 0:15:7.310 Nathan Lasnoski We just wanna do them faster or easier or with higher customer sat. 0:15:7.320 –> 0:15:9.340 Nathan Lasnoski We’re not changing the business per se. 0:15:9.410 –> 0:15:36.460 Nathan Lasnoski We’re driving efficiency within the business and then there’s this other side, which is the disruptive innovation and disruptive innovation is essentially, if I had to start my business over again, how would I bring to market my business with AI and sometimes the hardest organizations disrupt is yourself, especially when your organization is successful because you’re making money and you’re reacting in threats and you’re incrementally improving your product. 0:15:37.270 –> 0:15:59.70 Nathan Lasnoski But on the side comes this other organization that maybe offers a product that’s less feature rich at a lower price or a more incremental value stream and enables a set of customers to onboard that wouldn’t have done business with you because of where you are in the market or starts to pull customers away from you because of the way they’re going to mark it. 0:15:59.310 –> 0:16:9.590 Nathan Lasnoski So I’m working with a customer that is bringing to market the ability to create financial packages with customers that alternatively would have required a loan officer. 0:16:9.890 –> 0:16:15.730 Nathan Lasnoski Now, not requiring that loan officer, they’re able to take 2 to $3000 off of that process. 0:16:16.100 –> 0:16:21.910 Nathan Lasnoski This ability for them to engage in the market and completely disrupt the existing business model that exists. 0:16:23.750 –> 0:16:31.870 Nathan Lasnoski So this is the kind of value analysis that you should look to create as you go through the executive alignment and your envisioning sessions. 0:16:32.10 –> 0:16:41.940 Nathan Lasnoski You want to generate a value analysis where you’re able to clearly understand the categories value descriptions is an operational savings or revenue production. 0:16:41.950 –> 0:16:43.950 Nathan Lasnoski How hard is it to do like? 0:16:43.990 –> 0:16:46.100 Nathan Lasnoski Sometimes you’re moonshot idea. 0:16:46.160 –> 0:16:51.670 Nathan Lasnoski That’s not ready yet that you’re going to put investment behind isn’t the first thing you’re going to focus on because it’s not ready. 0:16:51.680 –> 0:16:52.600 Nathan Lasnoski Maybe the data is not ready. 0:16:53.380 –> 0:16:56.430 Nathan Lasnoski Ran ran a contract review process. 0:16:56.440 –> 0:16:57.510 Nathan Lasnoski Could be done tomorrow. 0:16:57.580 –> 0:17:13.290 Nathan Lasnoski It’s an easy use case to go after, so I encourage every organization to be very prescriptive about where you put your eggs and what baskets and focus energy around getting those ideas that get cut down into your focus list. 0:17:13.380 –> 0:17:15.320 Nathan Lasnoski Broaden the tent of engaging your teams. 0:17:16.790 –> 0:17:21.830 Nathan Lasnoski So let’s talk a little bit about use cases that organizations are taking advantage of. 0:17:22.540 –> 0:17:27.10 Nathan Lasnoski So on the left hand side here you can see this customer service use case. 0:17:27.20 –> 0:17:35.610 Nathan Lasnoski This is probably the most popular use case of them all, and the reason why is because we have a lot of opportunity to leverage data. 0:17:35.620 –> 0:17:52.140 Nathan Lasnoski We already have manuals or customer support data or knowledge based data to be able to answer questions from customers or enable customer support teams to be able to provide faster return on resolution to our customers. 0:17:52.720 –> 0:17:57.70 Nathan Lasnoski The less time than our customers sit and unhappy states, the better. 0:17:58.60 –> 0:18:0.560 Nathan Lasnoski How can I improve their experience of my products? 0:18:1.760 –> 0:18:4.350 Nathan Lasnoski The next use case is things like order status. 0:18:4.360 –> 0:18:5.530 Nathan Lasnoski Where’s my thing? 0:18:5.540 –> 0:18:6.990 Nathan Lasnoski When is it being delivered? 0:18:7.40 –> 0:18:13.130 Nathan Lasnoski Tends to be very, very popular, especially predicting and providing better data to your customers. 0:18:13.140 –> 0:18:32.0 Nathan Lasnoski You set better expectations, sales, quoting acceleration, any scenario where I can reduce the amount of time a customer needs or a number of steps they need to take or information they don’t have between their interest and being able to transact with me is a win. 0:18:32.330 –> 0:18:40.40 Nathan Lasnoski How can I shorten that cycle and drive up accuracy so when they do transact they get the best possible outcome? 0:18:40.410 –> 0:18:53.740 Nathan Lasnoski This is an area of dramatic revenue increase because if I can out compete my customer, my competitors on time or price or the approach that I’m using with information, I’m going to win more deals. 0:18:55.360 –> 0:19:2.690 Nathan Lasnoski Data mining, the ability to ask questions of business systems and provide information back and then this scenario here process automation. 0:19:2.920 –> 0:19:4.30 Nathan Lasnoski Anything you do, a lot. 0:19:4.100 –> 0:19:11.490 Nathan Lasnoski Invoices Peel review running a financial process of approval, preparing notes from sales calls. 0:19:12.180 –> 0:19:22.60 Nathan Lasnoski Any scenario where you’re having to do something over and over again or as a normal part of your business process, generative AI has the capability to help. 0:19:23.30 –> 0:19:34.220 Nathan Lasnoski So this all fits into a broader set of use cases that I’m not going to go into every single one of these, just that know that generative use case matches up to situations that we’ve been doing forever. 0:19:34.310 –> 0:19:37.390 Nathan Lasnoski So over here you’ve seen things like supply chain and predictive maintenance. 0:19:38.100 –> 0:19:50.130 Nathan Lasnoski Those are some of the first AI projects we did 8 some years ago that are generating revenue, operation, revenue savings or operational savings and revenue production right now because they are. 0:19:50.280 –> 0:20:0.770 Nathan Lasnoski They’re old use cases, and what’s you’re finding for many of those use cases is all the chips they had to put on the table in order to pursue those, say 7-8 years ago. 0:20:0.980 –> 0:20:6.40 Nathan Lasnoski You don’t have to put as many chips on the table anymore, because much of that has been automated. 0:20:6.380 –> 0:20:17.930 Nathan Lasnoski So we’re getting to a point where many of these use cases you’re seeing on the table here are opportunities for us to get involved and get engaged without as much investment in order to get started and to get outcome. 0:20:19.890 –> 0:20:26.360 Nathan Lasnoski So before I get into showing you some of those examples I wanna hit on some of the concerns that many organizations have as they get started. 0:20:27.150 –> 0:20:29.780 Nathan Lasnoski The first concern is data privacy. 0:20:29.950 –> 0:20:32.760 Nathan Lasnoski Everybody has this sort of fear, uncertainty and doubt. 0:20:32.830 –> 0:20:47.980 Nathan Lasnoski Regarding what about data privacy and what happens to my data once an AI system starts taking starts using that and doing something with it the the thing I want you to remember is that any business system that’s leveraging artificial intelligence is gonna use private instances of these platforms. 0:20:47.990 –> 0:20:58.420 Nathan Lasnoski Private instances of chat, GPT models, private instances of machine learning models used for things like supply chain, private instances for analysis of images. 0:20:58.570 –> 0:21:0.0 Nathan Lasnoski These are all private instances. 0:21:0.10 –> 0:21:4.980 Nathan Lasnoski They’re not data that’s shared with anyone else besides you and your business users that need to use it. 0:21:5.530 –> 0:21:15.820 Nathan Lasnoski Also being able to filter that data to the right people so the right people see their right data and don’t see the other they they shouldn’t see is important construct of just building a successful AI platform. 0:21:17.460 –> 0:21:25.150 Nathan Lasnoski This next scenario is data not being ready so often I see organizations say we’d love to do AI, but our data isn’t ready. 0:21:25.380 –> 0:21:28.690 Nathan Lasnoski Well, data readiness isn’t a broad concept. 0:21:28.700 –> 0:21:30.390 Nathan Lasnoski It’s a focused concept. 0:21:30.400 –> 0:21:43.350 Nathan Lasnoski It’s an understanding of is data ready for my use case and how can I take a revenue production or operational savings oriented view of how I enable data readiness for my organization in. 0:21:43.360 –> 0:21:49.20 Nathan Lasnoski This allows you to take a use case first approach that’s tied to value rather than just like. 0:21:49.30 –> 0:21:54.600 Nathan Lasnoski I’m gonna lift all the data and then then I can go to the ball like then I’ll be able to have data value from it. 0:21:54.670 –> 0:21:55.160 Nathan Lasnoski No. 0:21:55.250 –> 0:22:5.120 Nathan Lasnoski Focus on how I drive value from data by not only showing data, but getting prescriptive with data and enabling it to be able to trive actual outcomes. 0:22:6.760 –> 0:22:9.470 Nathan Lasnoski This middle use case we talked about human displacement. 0:22:9.940 –> 0:22:11.470 Nathan Lasnoski People are concerned you’ll get. 0:22:11.700 –> 0:22:15.230 Nathan Lasnoski You’ll start talking to our official intelligence and they think, how does this impact me? 0:22:15.310 –> 0:22:16.940 Nathan Lasnoski How does this impact my job? 0:22:17.10 –> 0:22:20.960 Nathan Lasnoski What is the future of my job in the context of leveraging artificial intelligence? 0:22:21.530 –> 0:22:22.550 Nathan Lasnoski Get in front of this. 0:22:23.440 –> 0:22:35.30 Nathan Lasnoski The best leaders I see in AI are the ones that bring their team together and enable them to see the future that’s going to exist as a result of AI by helping them to be able to drive value from it. 0:22:35.260 –> 0:22:41.330 Nathan Lasnoski Human displacement is about human enablement and enabling its switch around to make them part of the solution. 0:22:41.340 –> 0:22:46.310 Nathan Lasnoski Broadening the tent of who’s involved a hallucination, one of the bigger? 0:22:46.350 –> 0:22:56.790 Nathan Lasnoski Probably the second biggest topic is lunation, the idea that as I build one of these systems, how do we make sure that what it provides back to my customers or to my internal users as accurate? 0:22:57.260 –> 0:23:2.990 Nathan Lasnoski This is where not only are you able to start driving those advancements around. 0:23:3.0 –> 0:23:9.890 Nathan Lasnoski Quality yourself and your AI models, but oftentimes, like we’re not measuring the processes as they stand, like I don’t even know if the humans are doing it right. 0:23:10.140 –> 0:23:15.160 Nathan Lasnoski And this is a great place for us to be able to evaluate are we providing the right answers to customers? 0:23:15.530 –> 0:23:21.850 Nathan Lasnoski What is our accuracy is our is our production forecast accurate? 0:23:21.860 –> 0:23:31.330 Nathan Lasnoski What to what extent is it accurate today and how do we compare that against how we can improve it in the future and then very closely tied to quality is bias. 0:23:31.580 –> 0:23:41.710 Nathan Lasnoski How do I ensure that as I build an AI system that I’m barely engaging the data to provide accurate results? 0:23:41.800 –> 0:23:49.150 Nathan Lasnoski I’m I enabling good outcomes as results of that data are the outcomes that I’m enabling ones that are enabled by correct data? 0:23:49.280 –> 0:23:51.710 Nathan Lasnoski Am I using good data to be able to produce that AI model? 0:23:51.920 –> 0:23:56.950 Nathan Lasnoski This all factors into the picture as well and is super important on any use case that you pursue. 0:23:58.340 –> 0:24:1.30 Nathan Lasnoski Alright, so let’s talk a little bit about some examples. 0:24:1.40 –> 0:24:8.630 Nathan Lasnoski So I’ll walk you through a whole bunch of different types of examples that customers are using, so I’ve been using this example for a while. 0:24:8.640 –> 0:24:11.720 Nathan Lasnoski I think it’s a great starter example because it’s very tangible. 0:24:12.130 –> 0:24:15.660 Nathan Lasnoski This idea of surfacing information from documents. 0:24:16.10 –> 0:24:23.580 Nathan Lasnoski This is an example of a boat that is being surfaced for customer service purposes and for sales purposes. 0:24:23.590 –> 0:24:27.20 Nathan Lasnoski So this is a Brunswick sundancer 370. 0:24:27.90 –> 0:24:30.580 Nathan Lasnoski You can see that I’m asking you a question about the SUNDANCER 370. 0:24:30.590 –> 0:24:42.690 Nathan Lasnoski It’s providing me a very reasonable response and it’s giving me a citation as to where it got that data and providing me the ability to be able to see that information on the right hand side. 0:24:42.760 –> 0:24:51.870 Nathan Lasnoski You can also see I’m asking about the draft or the maximum speed, and then here you can see where I’m getting that data within the actual document and servicing it back to the customers. 0:24:51.940 –> 0:24:53.170 Nathan Lasnoski Why is this so important? 0:24:53.180 –> 0:24:59.390 Nathan Lasnoski Well, I want a customer to have a great onboard experience and they don’t have a great onboard experience. 0:24:59.440 –> 0:25:2.970 Nathan Lasnoski If something goes wrong or they don’t know the right information about the boat. 0:25:3.320 –> 0:25:20.490 Nathan Lasnoski So for example, I happen to have one, not this specific boat, but I have a AC ray boat and it had a failure occur to it and I had to try to fix it and I spent four hours just trying to figure out what kind of lubricant I had to put in to solve for this light that came out in the boat. 0:25:20.640 –> 0:25:31.550 Nathan Lasnoski And this is the case for many many use cases with manufacturers or with customer service scenarios where you go out to the site and you’re like I can’t make heads or tails of this like, this is what boat do I have? 0:25:31.560 –> 0:25:36.110 Nathan Lasnoski There’s like 62 different boats listed here, like how do I even know what I bought five years ago? 0:25:36.460 –> 0:25:42.540 Nathan Lasnoski This is an opportunity for us to be able to ease the customer through the process of being able to answer questions that they have. 0:25:43.700 –> 0:25:46.230 Nathan Lasnoski This is another customer service anaro for Generac. 0:25:46.280 –> 0:25:49.210 Nathan Lasnoski So Generac has a internal chat bot. 0:25:49.220 –> 0:25:53.270 Nathan Lasnoski It’s goal is to answer questions about standby home generators. 0:25:53.400 –> 0:25:56.790 Nathan Lasnoski This is a question being asked about that generator. 0:25:56.900 –> 0:26:8.440 Nathan Lasnoski This is the AI response associated with it and you can see I’m pointing to a source I’m giving sort of question a ability for me to ask how are you? 0:26:8.450 –> 0:26:9.420 Nathan Lasnoski How’s this going for you? 0:26:9.430 –> 0:26:10.620 Nathan Lasnoski Like is this answering the question? 0:26:10.630 –> 0:26:20.360 Nathan Lasnoski Is it not answering the question and the results of something like this is I can enable customer service teams to be able to answer questions faster. 0:26:20.890 –> 0:26:27.520 Nathan Lasnoski Don’t you always hate it when, like, you call in and you get put on hold like I always hate that myself like man, like I just hate being put on hold. 0:26:27.670 –> 0:26:30.380 Nathan Lasnoski Can I enable that individual to not get put on hold? 0:26:30.390 –> 0:26:34.120 Nathan Lasnoski Can I enable that question to be answered more quickly and with greater accuracy? 0:26:34.130 –> 0:26:37.830 Nathan Lasnoski And then measure over time, is that question answered accurately? 0:26:37.940 –> 0:26:40.130 Nathan Lasnoski So really cool use case. 0:26:40.600 –> 0:26:46.350 Nathan Lasnoski Something that’s gonna provide a ton of value and eventually get put right in front of the customers so they can answer their own questions. 0:26:48.40 –> 0:26:48.710 Nathan Lasnoski This is no. 0:26:48.720 –> 0:26:50.230 Nathan Lasnoski This is where it starts to get really cool. 0:26:50.760 –> 0:26:57.60 Nathan Lasnoski So once you can see here is I’m starting to do things in a multimodal way. 0:26:57.70 –> 0:27:2.370 Nathan Lasnoski So this is an instruction catalog for an IKEA shelf. 0:27:2.700 –> 0:27:18.160 Nathan Lasnoski I’ve spent many long hours putting together IKEA stuff and I can say that the instruction manuals are pretty good actually sometimes, but it’s something that AI would not have been particularly good at diagnosing or understanding. 0:27:18.560 –> 0:27:23.460 Nathan Lasnoski The reason why is because I would need to say let’s train it on one a hammer is. 0:27:23.470 –> 0:27:25.120 Nathan Lasnoski Let’s train it on what a screw is. 0:27:25.130 –> 0:27:29.530 Nathan Lasnoski Let’s train it on what a shelf piece is and how that this is it being put together. 0:27:29.540 –> 0:27:31.810 Nathan Lasnoski And like, here’s this another piece. 0:27:31.820 –> 0:27:34.540 Nathan Lasnoski Here we’d have to train it on all those sub elements. 0:27:34.690 –> 0:27:58.170 Nathan Lasnoski So how do I get value from this just to I think 2 days ago Microsoft GA or not G8 released for use the chat sheet TV capabilities it would GBTV does is enables the model to be able to understand and inspect information that exists within visual senses, not just text. 0:27:58.180 –> 0:28:7.430 Nathan Lasnoski So many of the manuals they can for that boat, for example, and there’s a manual that has an image of the cockpit and an image of the cockpit, has a peon. 0:28:7.440 –> 0:28:12.130 Nathan Lasnoski It ABCDE and it’s pointing to different things and you say like where is the throttle? 0:28:12.180 –> 0:28:18.150 Nathan Lasnoski Well, just the creating an AI model in that in the past is really complicated because you had to try on all those pieces. 0:28:18.300 –> 0:28:20.340 Nathan Lasnoski Now I can bring that to market faster. 0:28:20.350 –> 0:28:21.490 Nathan Lasnoski You can see right here. 0:28:21.580 –> 0:28:23.410 Nathan Lasnoski It knows what’s happening in step seven. 0:28:23.420 –> 0:28:24.230 Nathan Lasnoski I’m asking the question. 0:28:24.570 –> 0:28:26.240 Nathan Lasnoski It’s doing that correctly. 0:28:26.250 –> 0:28:35.250 Nathan Lasnoski It’s providing me what’s happening in step seven in a text based format, so this is an example that being used for defect detection. 0:28:35.520 –> 0:28:36.850 Nathan Lasnoski So this is a. 0:28:37.860 –> 0:28:39.280 Nathan Lasnoski This is a successful screw. 0:28:39.290 –> 0:28:46.190 Nathan Lasnoski We’re giving you an example of successful like this is what a screw should look like as it comes off of our supply chain. 0:28:46.200 –> 0:28:48.960 Nathan Lasnoski Right, here’s a bunch of other screws. 0:28:50.960 –> 0:28:58.690 Nathan Lasnoski Which of these are good screws so you can see here the screw head within the green circle is a standard for non defective piece. 0:28:58.700 –> 0:28:59.470 Nathan Lasnoski I will compare it. 0:29:0.60 –> 0:29:1.930 Nathan Lasnoski Number one is defective. 0:29:2.180 –> 0:29:4.130 Nathan Lasnoski #2 is not defective. 0:29:4.190 –> 0:29:6.630 Nathan Lasnoski So 1 defective, 2 not defective. 0:29:6.940 –> 0:29:16.220 Nathan Lasnoski What this is doing is it’s enabling us to be able to tell it what good looks like and then provide answers for what not good. 0:29:16.230 –> 0:29:28.310 Nathan Lasnoski Looks like very quickly providing value back to the organization to be able to do that inspection and this is something that we could we could do with traditional machine learning techniques but not as fast. 0:29:28.320 –> 0:29:32.190 Nathan Lasnoski Now we can do it much faster and provide much more input. 0:29:32.340 –> 0:29:33.830 Nathan Lasnoski Now here’s where it gets really cool. 0:29:33.840 –> 0:29:34.610 Nathan Lasnoski Check this out. 0:29:34.650 –> 0:29:40.130 Nathan Lasnoski OK, so this is a video of a guy walking around a car magine that you got in an accident. 0:29:40.140 –> 0:29:44.600 Nathan Lasnoski OK, I know that’s not super fun, but like, imagine something happened to your car and you’re walking. 0:29:44.610 –> 0:29:57.360 Nathan Lasnoski I hit a deer recently like it’s big nearby and walking around and taking a video of the car and then what happens is a result of that is you are having to write up as an insurance agent. 0:29:57.550 –> 0:29:59.170 Nathan Lasnoski Like what happened with that vehicle? 0:30:0.100 –> 0:30:7.420 Nathan Lasnoski So this is then description that was produced via AI of what happened with that image. 0:30:7.430 –> 0:30:14.890 Nathan Lasnoski That vehicle, the rear side of the blue Toyota Camry, has sustained significant damage characterized by deep scratches and scuff marks, blah blah blah blah blah, right? 0:30:16.0 –> 0:30:19.210 Nathan Lasnoski This enables insurance and enables repair shops. 0:30:19.220 –> 0:30:29.910 Nathan Lasnoski It enables the manufacturers to know how, how damaged, sustained in certain types of use cases think about vision in this context that like I don’t have to have a person create that draft. 0:30:29.920 –> 0:30:35.550 Nathan Lasnoski Maybe I have to have a person and review the draft and know what happened, but I don’t have to have a person create the draft. 0:30:35.560 –> 0:30:39.350 Nathan Lasnoski I can force multiply that individual and enable them to get more done. 0:30:39.420 –> 0:30:42.170 Nathan Lasnoski Or maybe just not have to do this particular part of the task. 0:30:42.180 –> 0:30:53.70 Nathan Lasnoski Maybe they can interact with the individual who sustained the damage, or the person can just take their own phone into that themselves and then have it input that as part of the app for video insurance. 0:30:53.580 –> 0:31:3.680 Nathan Lasnoski This is where GPTV gets really rad, like the ability to take this kind of content, they combine it then with the text based content to produce real outcomes. 0:31:4.50 –> 0:31:9.940 Nathan Lasnoski Is just amazing stuff and you can see how this is just going to keep driving and driving. 0:31:10.10 –> 0:31:15.760 Nathan Lasnoski Lack of a better word, more and more capabilities out of what AI can create within our businesses. 0:31:15.970 –> 0:31:18.960 Nathan Lasnoski So this sort of video engagement is going to be really cool. 0:31:20.190 –> 0:31:22.300 Nathan Lasnoski OK, so next use case. 0:31:22.390 –> 0:31:29.220 Nathan Lasnoski You know one thing I forgot to mention is as you are having additional questions, feel free to drop those things in the chat. 0:31:29.230 –> 0:31:38.720 Nathan Lasnoski I would love to answer other questions near the end or as we’re going as you have them all right, so this is an example of accelerating sales to close. 0:31:38.930 –> 0:31:51.610 Nathan Lasnoski So in this use case, what we’re doing is taking the natural language processing and we’re inspecting a question from a customer and then we’re taking that request to to quote something. 0:31:52.60 –> 0:31:56.610 Nathan Lasnoski We’re extracting the necessary information and then we are producing the quote. 0:31:56.620 –> 0:31:58.710 Nathan Lasnoski We’re accepting it or we’re modifying it. 0:31:58.800 –> 0:32:2.950 Nathan Lasnoski And then we’re turning that into the actual quote that gets sent back to the customer. 0:32:3.20 –> 0:32:5.210 Nathan Lasnoski Does that remove the human completely from that process? 0:32:5.220 –> 0:32:8.450 Nathan Lasnoski No, there’s still getting the prepared asset to review. 0:32:8.560 –> 0:32:11.130 Nathan Lasnoski They either modified or accept it and they send it back. 0:32:11.220 –> 0:32:14.800 Nathan Lasnoski You could potentially take a subset of that with the customer self quote. 0:32:14.890 –> 0:32:18.870 Nathan Lasnoski In that case, the customers reviewing the quote themselves gets sent forward. 0:32:18.950 –> 0:32:21.380 Nathan Lasnoski What we’re doing here is we’re deducing the time to quote. 0:32:21.470 –> 0:32:26.600 Nathan Lasnoski So in this case, we’re taking time to quote for this customer from hours down to minutes. 0:32:26.910 –> 0:32:29.440 Nathan Lasnoski Just think about the aggregate time savings of that. 0:32:29.450 –> 0:32:38.740 Nathan Lasnoski In this case, the customer has 600 sales reps, 600 sales reps brought the coolest they do from hours down to minutes like that is ridiculous. 0:32:38.750 –> 0:32:41.550 Nathan Lasnoski Kind of improvement from their ability to engage their customers. 0:32:42.420 –> 0:32:51.160 Nathan Lasnoski This is also being used for use cases where you’re trying to optimize price, so I have a customer in working with where they have a set of price they use on the external website. 0:32:51.170 –> 0:32:59.750 Nathan Lasnoski They have a set of price they use on their internal users when they quote and they’re trying to normalize those across how they are delivering quotes to their customers. 0:33:0.60 –> 0:33:9.610 Nathan Lasnoski They can use AI to be able to optimize that quoting process to be able to provide the best price for the opportunity for the customer and for themselves. 0:33:9.620 –> 0:33:14.640 Nathan Lasnoski So they gain the right margin, so sales quoting is a a great opportunity within accounts. 0:33:14.650 –> 0:33:36.80 Nathan Lasnoski You can also take this into complex order flows, ones where you’re delivering a CAD drawing for example like save that this requires a CAD drawing and in order to do something with it I need to inspect that CAD and know the components of it, know if it’s put together properly, know if the measurements are right, understand where there’s flaws within the way that they’ve designed it. 0:33:36.330 –> 0:33:41.250 Nathan Lasnoski These are areas where we can apply AI as well to be able to provide the right kinds of outcomes. 0:33:42.80 –> 0:33:48.0 Nathan Lasnoski What get ultimately getting to points where you might be doing a RR experiences to help someone generate the right? 0:33:48.740 –> 0:33:49.750 Nathan Lasnoski No, the right platform. 0:33:49.760 –> 0:33:56.770 Nathan Lasnoski So this is the window manufacturer generating a AR window and saying here fits inside of this spot. 0:33:56.920 –> 0:33:59.370 Nathan Lasnoski What’s the right window that would fit here? 0:33:59.560 –> 0:34:8.770 Nathan Lasnoski Scrum out all the things not appropriate for my location or region, and let’s walk through the example of how I could build a better experience for that customer. 0:34:8.960 –> 0:34:14.170 Nathan Lasnoski What’s necessary to solve for this is the schema that exists underneath it. 0:34:14.180 –> 0:34:18.60 Nathan Lasnoski So sometimes when people say data is not right, this is a good example where you really the data ready right? 0:34:18.70 –> 0:34:27.840 Nathan Lasnoski You need the data for the schema to be able to serve with the customers going to be able to do from a AR experience and then ultimately produce the install guide that’s being used by the customer. 0:34:28.110 –> 0:34:31.910 Nathan Lasnoski So there’s opportunities for really game changing things. 0:34:31.920 –> 0:34:39.250 Nathan Lasnoski In this case, you’re really getting around the idea of the guy going to Home Depot and talking with the person who pulls out the book and like walks through in the book. 0:34:39.260 –> 0:34:41.380 Nathan Lasnoski What they’re going to select and doesn’t really know. 0:34:41.390 –> 0:34:44.500 Nathan Lasnoski And then calls a second guy to your house to do the measurements and stuff. 0:34:44.550 –> 0:34:45.820 Nathan Lasnoski It’s like a terrible experience. 0:34:46.250 –> 0:34:52.310 Nathan Lasnoski Imagine if I was doing this all in a self quote and I could visualize it right in the comfort of my own home. 0:34:52.400 –> 0:34:55.590 Nathan Lasnoski This is where AI is enabling some really game changing experiences. 0:34:55.600 –> 0:34:57.210 Nathan Lasnoski Using text and video. 0:34:58.850 –> 0:35:1.910 Nathan Lasnoski Alright, so let’s move into the supply chain use cases. 0:35:1.920 –> 0:35:8.250 Nathan Lasnoski Supply chain is one of the older but more valuable opportunities within your organization. 0:35:9.910 –> 0:35:27.840 Nathan Lasnoski So this example is from 2018 and I reason I say that is because I want you to know this is not a new concept getting value from your supply chain and applying AI to it is not a new concept, but it is a not a concept that the cost of engaging continues to be reduced. 0:35:28.130 –> 0:35:32.100 Nathan Lasnoski So in this case this is a $2 billion organization. 0:35:32.330 –> 0:35:42.520 Nathan Lasnoski They remanufacture cell phones and encourages, and they freed up 50 to $80 million of capital yearly as a component of optimizing their inventory spend. 0:35:43.360 –> 0:36:7.450 Nathan Lasnoski They’re essentially the relationship between demand and inventory they have on hand to be able to produce their products and why this was so impactful for them is they were able to do a pre COVID during COVID and after COVID and be able to react to the different types of changes happening and be able to understand and story tell what happens if these things occur and what would it choices. 0:36:7.460 –> 0:36:9.10 Nathan Lasnoski Would we be making based upon that? 0:36:9.220 –> 0:36:20.780 Nathan Lasnoski What made this even more interesting is that as they went down this path, they were able to start predicting the outcomes for their customers and for their suppliers. 0:36:20.830 –> 0:36:33.420 Nathan Lasnoski So they essentially became like the demand inventory forecasting for their customer to make them stickier with that organization to help them optimize their own ordering and to build a stronger relationship. 0:36:33.510 –> 0:36:35.770 Nathan Lasnoski And they ultimately were able to start charging for that. 0:36:35.870 –> 0:36:42.610 Nathan Lasnoski So like this becomes almost a profit Center for an organization to be able to engage in the way that you become sticky with a customer. 0:36:43.550 –> 0:36:54.740 Nathan Lasnoski So a lot of interesting stuff here, if you will send over the deck, there’s a great video about this use case, but it’s a fantastic opportunity for an organization to get value from AI and yours. 0:36:54.800 –> 0:37:0.740 Nathan Lasnoski Now one thing I would call out on this, I just want to be built this many ML use cases. 0:37:2.430 –> 0:37:7.260 Nathan Lasnoski Organizations are still banging away on Excel spreadsheets like they’ve got in your piece system. 0:37:7.570 –> 0:37:12.200 Nathan Lasnoski We’re kind of regardless of what it is, dynamics, SAP, blah, blah, blah, right. 0:37:12.280 –> 0:37:16.460 Nathan Lasnoski There’s still like, taking that stuff, exporting it to excel, monkeying with it. 0:37:16.530 –> 0:37:20.70 Nathan Lasnoski It’s now no longer the most recent data, and then they’re making choices. 0:37:21.110 –> 0:37:23.640 Nathan Lasnoski Some companies they’ve gone on tried SAS providers. 0:37:23.650 –> 0:37:32.680 Nathan Lasnoski My experience with the SAS providers space has been you tend to put as much energy in with the SAS provider as you would just building your own and getting competitive advantage. 0:37:32.990 –> 0:37:47.930 Nathan Lasnoski So my suggestion is get competitive advantage drive opportunity around machine learnings in the supply chain space where you can truly create differentiation from your competitors and where your models really know you. 0:37:48.230 –> 0:37:56.120 Nathan Lasnoski And this is something that it’s really key to this, these truly improved your business, if it knows you and this is where you can gain that opportunity. 0:37:58.150 –> 0:38:3.300 Nathan Lasnoski OK, let’s get into the prediction space prediction is like let’s actually anything that you’re doing. 0:38:3.350 –> 0:38:6.800 Nathan Lasnoski So this is an example of a company that’s over in Waukesha. 0:38:6.810 –> 0:38:13.860 Nathan Lasnoski Her as electronics and what they do is essentially melt metal and turn it into composites. 0:38:14.70 –> 0:38:23.360 Nathan Lasnoski So if everybody remembers Terminator 2 right, like in Terminator 2, Arnold’s being lowered into the big VAT of molten metal, like same kind of idea, right? 0:38:23.370 –> 0:38:27.550 Nathan Lasnoski Except they need to measure it and know if they’re going to have a problem with that production process. 0:38:27.920 –> 0:38:30.390 Nathan Lasnoski And if they do have a problem, they have to throw it out. 0:38:30.400 –> 0:38:31.990 Nathan Lasnoski And that’s a very expensive error. 0:38:32.320 –> 0:38:38.690 Nathan Lasnoski So what this AI is doing is helping them to predict how they produce that on the floor product. 0:38:38.700 –> 0:38:42.850 Nathan Lasnoski And when I think this is really interesting is it’s an example of the relationship between it and OT. 0:38:43.420 –> 0:39:8.750 Nathan Lasnoski So a lot of times these OT platforms, they’re like siloed walled gardens in many cases, because the things running on XP in this case, we built A1 way interface to that out environment to enable them to be able to get that information and allow them to be able to make accurate choices about how they’re they’re running their production process by integrating with their OTT network, create opportunity for them to be able to gain improvement. 0:39:9.890 –> 0:39:16.360 Nathan Lasnoski So this is an example of prediction impacting your business process, but then the next example takes it up another notch. 0:39:16.450 –> 0:39:20.60 Nathan Lasnoski This is an example of a prediction impacting prenatal care. 0:39:20.470 –> 0:39:24.420 Nathan Lasnoski So this is a company called Parrigin and what they do is they track fetal. 0:39:24.470 –> 0:39:26.620 Nathan Lasnoski A lot of things, but they track fetal heart rate. 0:39:26.890 –> 0:39:29.560 Nathan Lasnoski So like, I don’t know if anyone has had a baby recently. 0:39:29.570 –> 0:39:30.340 Nathan Lasnoski I had one. 0:39:30.550 –> 0:39:31.360 Nathan Lasnoski I didn’t have one. 0:39:31.370 –> 0:39:35.330 Nathan Lasnoski My wife had one about four months ago and what you have in the. 0:39:37.460 –> 0:39:45.190 Nathan Lasnoski The the room where you’re having the baby is they’re slow at this little ticker tape comes off and it’s doing the fetal heart rate and the heart rate over the mother. 0:39:45.500 –> 0:39:58.10 Nathan Lasnoski And what this does is it digitizes it and then provides an interpretation of what’s going to happen in the future in alignment with what it thinks is going to happen and why I think this is really important is people always like, what about hallucinations? 0:39:58.20 –> 0:39:58.590 Nathan Lasnoski What about? 0:39:59.140 –> 0:40:0.510 Nathan Lasnoski What about the quality? 0:40:0.520 –> 0:40:0.950 Nathan Lasnoski And so on. 0:40:1.120 –> 0:40:13.370 Nathan Lasnoski Well, this is where like if you get this wrong and there’s negative outcomes to a person and this is where we’re using capabilities like this in conjunction with the doctor to be able to provide. 0:40:13.640 –> 0:40:18.90 Nathan Lasnoski Very excellent outcomes that wouldn’t have been wouldn’t have existed before with a human. 0:40:18.100 –> 0:40:33.150 Nathan Lasnoski Just looking at what happened in the past and trying to interpret that themselves, so many opportunities that are provided and this is enabling this to be to be able to be engaged not just for the mother, but also up to four babies that would exist inside of the womb. 0:40:33.160 –> 0:40:36.610 Nathan Lasnoski So really awesome capabilities result of AI. 0:40:38.180 –> 0:40:38.570 Nathan Lasnoski Alright. 0:40:38.580 –> 0:40:41.910 Nathan Lasnoski And then something very similar to this, I I want everybody to put in the chat. 0:40:41.920 –> 0:40:46.300 Nathan Lasnoski What do you think is in this video? 0:40:46.310 –> 0:40:47.580 Nathan Lasnoski What is this an image of? 0:40:47.930 –> 0:40:48.820 Nathan Lasnoski I’ll give you a second. 0:40:48.870 –> 0:40:49.970 Nathan Lasnoski What is this an image of? 0:41:7.130 –> 0:41:8.660 Nathan Lasnoski Bricks, not bricks. 0:41:15.630 –> 0:41:21.730 Nathan Lasnoski OK, so this is I have one person say it’s a colon, it’s not a colon. 0:41:22.480 –> 0:41:23.350 Nathan Lasnoski This is the. 0:41:23.520 –> 0:41:26.110 Nathan Lasnoski This is the inside of a pipe. 0:41:28.320 –> 0:41:30.350 Nathan Lasnoski That’s the image, by the way, talking. 0:41:30.780 –> 0:42:0.520 Nathan Lasnoski This is the inside of a pipe and what this company does is it goes inside of the pipe and it looks for errors within that pipe are not errors like cracks and other problems like one of the things that I didn’t know until we engaged this organization was how much water waste there is like both from an output and input stance, and how much these cracks create problems for the businesses, the fields or whatever that exists around these pipes. 0:42:1.390 –> 0:42:6.170 Nathan Lasnoski So what this company does is they go down in the pipes and they look for these problems and then they correct them. 0:42:7.110 –> 0:42:14.140 Nathan Lasnoski And I’m I’ve never done this, but I have watched oshinko redemption, and I know that that’s crawling inside of one of those pipes would be really awful. 0:42:14.410 –> 0:42:20.980 Nathan Lasnoski So what this does is it goes inside the pipe and it looks for those kinds of errors and it finds them and then it says issue found right? 0:42:21.40 –> 0:42:32.630 Nathan Lasnoski That indicates where those problems are and where they should focus their energy, and why this is so important is because it provides the opportunity for an organization to be able to optimize the work done by the human. 0:42:33.40 –> 0:42:35.930 Nathan Lasnoski Humans aren’t really good at just like watching continuous video, right? 0:42:35.940 –> 0:42:38.20 Nathan Lasnoski They get tired, they blink, they’re not focused. 0:42:38.260 –> 0:42:51.640 Nathan Lasnoski This is an ability for us to then to look at specific things and do something about it, and I think it’s really kind of game changing idea and this gets even better as you start enabling GPB versus kind of historical training techniques. 0:42:54.110 –> 0:42:55.820 Nathan Lasnoski OK, let’s talk about customer intelligence. 0:42:56.680 –> 0:42:58.230 Nathan Lasnoski 2 examples I want to hit on here. 0:42:58.640 –> 0:43:1.490 Nathan Lasnoski So the first example is about where’s my order. 0:43:1.900 –> 0:43:13.70 Nathan Lasnoski So in this example you can see we have green on our delivery cycle, but we have this yellow section that represents an unknown and then we also have this section where we don’t have data at all. 0:43:13.860 –> 0:43:18.670 Nathan Lasnoski Oftentimes, humans look at this and there’s an upset customer, and the customer says, well, where is my order? 0:43:18.680 –> 0:43:22.780 Nathan Lasnoski When is it getting to me so the person runs around this individual? 0:43:22.790 –> 0:43:23.510 Nathan Lasnoski Who does this job? 0:43:23.520 –> 0:43:27.190 Nathan Lasnoski She told me she’s her job is to be the human connector of digital systems. 0:43:27.720 –> 0:43:30.190 Nathan Lasnoski And I’m like, wow, that’s very interesting that, that. 0:43:30.200 –> 0:43:43.820 Nathan Lasnoski And and I understand because these systems are not connected to each other, I can’t provide that data and she spends a lot of her time going asking questions of people as to when stuff’s going to be delivered and what they’re seeking to do. 0:43:43.830 –> 0:43:57.30 Nathan Lasnoski And is have been successful at is to start to predict what are the gaps within that delivery process and what are their delivery times based upon what we know using artificial intelligence to be able to provide that information back. 0:43:57.380 –> 0:44:10.420 Nathan Lasnoski So this goal of this platform is to say, how do I provide that data back and then doing so through a customer asynchronous experience where they can ask those questions of the supply chain and provide that answer back. 0:44:10.430 –> 0:44:13.520 Nathan Lasnoski So when the customer calls in and says, what’s the status of order X? 0:44:13.570 –> 0:44:24.450 Nathan Lasnoski It can provide you the answer to what the status of order X is, because we’ve done the leg work to be able to enable it to be able to tell the answer back and to be able to be better than not. 0:44:24.620 –> 0:44:25.640 Nathan Lasnoski Then partially what? 0:44:25.650 –> 0:44:32.960 Nathan Lasnoski The human could do on their own because what it enables is leveraging artificial intelligence to be able to do prediction against that supply chain. 0:44:33.900 –> 0:44:36.270 Nathan Lasnoski So your goal here is to set better expectations. 0:44:37.380 –> 0:44:47.700 Nathan Lasnoski The Domino’s Pizza tracker as it stands in this use case is going to set better expectations because it’s able to bring all that data together and provide more outcome, more data back. 0:44:47.950 –> 0:44:50.340 Nathan Lasnoski Another thing I’m gonna just keep this up for a second. 0:44:50.350 –> 0:44:53.740 Nathan Lasnoski Another thing, I don’t have an image of, but I think is really interesting. 0:44:53.750 –> 0:45:0.650 Nathan Lasnoski Example is where data can be surfaced to provide more information back as a service as a value. 0:45:0.660 –> 0:45:2.100 Nathan Lasnoski So I have a company I work with. 0:45:2.390 –> 0:45:9.330 Nathan Lasnoski They’re a $50 billion food distributor and they make all their money by distributing food extremely low margin. 0:45:9.620 –> 0:45:15.230 Nathan Lasnoski So we’ll engage customers and deliver low margin food distribution to those those companies. 0:45:16.20 –> 0:45:28.370 Nathan Lasnoski But what you probably also know is that the restaurant business is extremely difficult to do work to do business apps right to the rate of restaurants that fail is like 9 out of 10. 0:45:28.700 –> 0:45:33.330 Nathan Lasnoski So what this company knows is what are the successful restaurants doing? 0:45:33.560 –> 0:45:34.690 Nathan Lasnoski What are they buying at? 0:45:34.700 –> 0:45:36.330 Nathan Lasnoski What kinds of products are they using? 0:45:36.480 –> 0:45:37.570 Nathan Lasnoski How often do they buy? 0:45:37.580 –> 0:45:38.640 Nathan Lasnoski What price do they buy in it? 0:45:38.650 –> 0:45:39.560 Nathan Lasnoski What are they selling at? 0:45:39.970 –> 0:45:45.940 Nathan Lasnoski And they can start to understand what a successful restaurant looks like versus what a non successful restaurant looks like. 0:45:46.30 –> 0:45:53.80 Nathan Lasnoski So they can actually sell back that information to their customer to help them to be able to be more successful. 0:45:53.120 –> 0:46:3.330 Nathan Lasnoski So less restaurants go out of business and essentially also creates a very sticky relationship with them and enables them to get into a higher margin business because they’re able to leverage the data. 0:46:3.340 –> 0:46:5.460 Nathan Lasnoski And this is really disruptive space is. 0:46:5.710 –> 0:46:10.800 Nathan Lasnoski How do I create an entirely new way of doing business with my customers based upon what I know about them? 0:46:11.720 –> 0:46:23.450 Nathan Lasnoski And this is where you’re seeing not just like not just using data to provide like what Netflix shows, should I watch, but even more so to create an entirely new understanding of how I provide recommendations back. 0:46:23.460 –> 0:46:27.210 Nathan Lasnoski Same kind of idea we’re seeing in the financial services space about next specs to action. 0:46:27.560 –> 0:46:34.870 Nathan Lasnoski What actions should rather than me like calling my financial representative every quarter and saying, Oh yeah, cool looks OK? 0:46:34.920 –> 0:46:40.520 Nathan Lasnoski Like what if it was giving me more real time information about what I should be doing with My Portfolio? 0:46:40.640 –> 0:46:45.900 Nathan Lasnoski So I can be making better choices, so a lot of opportunity from the from the data stage. 0:46:46.390 –> 0:46:48.740 Nathan Lasnoski OK, let’s talk about process automation. 0:46:48.790 –> 0:46:54.150 Nathan Lasnoski Process automation is one of the richest opportunities because it’s just things that we do a lot. 0:46:54.640 –> 0:46:58.50 Nathan Lasnoski So any time in the process automation space, you’re doing a process. 0:46:58.60 –> 0:47:17.510 Nathan Lasnoski So this is customer contract automation, so customer contract automation is about analyzing and purchase order matching that sales order with a production order, validate it before it goes to production and then executing this is a step in a series of steps that a person has to go through to make sure that we’re producing the right thing especially in high value production. 0:47:17.960 –> 0:47:19.480 Nathan Lasnoski Each of those steps take time. 0:47:19.490 –> 0:47:25.990 Nathan Lasnoski Each of those steps realize something or find something like maybe I don’t have all the parts, or perhaps I ordered it wrong. 0:47:26.0 –> 0:47:29.830 Nathan Lasnoski Right, actually quoted it wrong, but the customer I’m building the wrong thing. 0:47:29.980 –> 0:47:33.820 Nathan Lasnoski All these steps are important to make sure we have better customer sat. 0:47:34.200 –> 0:47:39.740 Nathan Lasnoski So a lot of times with people have humans running through these processes, but a lot of times they don’t need to look at these processes. 0:47:39.750 –> 0:47:42.440 Nathan Lasnoski The each of these documents, because most of them are pretty cut and dry. 0:47:43.620 –> 0:48:0.830 Nathan Lasnoski So what you can do is have a I run that process, invalidate it, and then if there’s something that I can’t do, have an exception to the human intervention, have an exception where the human is getting a subset of those processes. 0:48:0.840 –> 0:48:4.940 Nathan Lasnoski Probably the most important ones for them to look at to take action about. 0:48:5.320 –> 0:48:23.20 Nathan Lasnoski So this is an example of taking information from text layout, pulling that information out into a markdown table, and in that being sent into extracted JSON that I can use to be able to automate any flavor of product. 0:48:23.110 –> 0:48:27.400 Nathan Lasnoski So this is taking it straight from an image into markdown into JSON. 0:48:28.330 –> 0:48:47.520 Nathan Lasnoski I can then also take that from images from other assets and automate it and where this has gotten really amazing is GPT 4 has taken the ability for us to do text analysis and turns that up to another notch where I couldn’t do it before because I hadn’t trained it on certain types of things. 0:48:47.770 –> 0:48:55.540 Nathan Lasnoski Now I’m able to further extend that to be able to provide more value back into the automation of the process. 0:48:55.690 –> 0:49:11.40 Nathan Lasnoski So this is huge like anything you just going through a list of things that you do often and finding opportunities where automation will let your organization scale rather than scaling by adding more people is a great opportunity for us to be able to create real outcomes. 0:49:11.940 –> 0:49:40.750 Nathan Lasnoski OK, want to show you the the last example that I’ve got here which is smart products and this is where you’re taking, you’re taking the thing that you create for your customers and you’re enabling it with AI and this is this is Brunswick smart boat smart voting platform, what they’re doing with the smart voting platform is they’re having a goal of being a leader in digital boating and by being a leader in digital boating, that means connected boat experience. 0:49:41.710 –> 0:49:43.700 Nathan Lasnoski The is the battery charged up. 0:49:43.990 –> 0:49:46.270 Nathan Lasnoski Is someone on my boat at my geofencing it? 0:49:46.280 –> 0:49:50.960 Nathan Lasnoski Well, it’s in the in the the the harbor, should I even be boating today? 0:49:50.970 –> 0:49:52.300 Nathan Lasnoski Is the weather gonna be great? 0:49:52.430 –> 0:50:2.820 Nathan Lasnoski Should I be taking this boat to Fayette or is are the waves going to be too significant for it, enabling me to have a great onboard experience building AI right within our product? 0:50:2.830 –> 0:50:16.640 Nathan Lasnoski You should earlier this week, Rockwell announced with Microsoft their factory talk platform has a new A copilot built-in to enable their users to be able to gain value from AI right within their factory talk platform. 0:50:16.710 –> 0:50:26.220 Nathan Lasnoski So if you build a product or you deliver a service, now is the time to be thinking about how does my product have AI injected into itself to deliver value? 0:50:26.430 –> 0:50:28.840 Nathan Lasnoski How does my customer get that value? 0:50:29.50 –> 0:50:32.900 Nathan Lasnoski And that’s where simply by having the product oriented mindset. 0:50:33.110 –> 0:50:42.380 Nathan Lasnoski By continuing to build more capability and enabling as a digital product, not simply a physical product or a person, product enables us to deliver more capabilities. 0:50:42.390 –> 0:50:47.410 Nathan Lasnoski A company that the product you’re building with, that new capability to deliver value. 0:50:52.140 –> 0:50:56.460 Nathan Lasnoski OK, so a couple of closing comments before you leave today. 0:50:56.930 –> 0:51:1.670 Nathan Lasnoski I want you to fill out the survey and the reason why I want you to do that is a couple fold. 0:51:1.680 –> 0:51:9.260 Nathan Lasnoski One, I want to get feedback and once continue to improve and provide more data back to you and help you to gain more value from AI. 0:51:9.980 –> 0:51:14.10 Nathan Lasnoski Umm, but then in addition to that, we want to help you gain that progress. 0:51:14.180 –> 0:51:21.390 Nathan Lasnoski So in that survey, you can sign up for us to help you bring this kind of presentation to your custom internal users. 0:51:21.460 –> 0:51:23.110 Nathan Lasnoski We can do art of the possible for you. 0:51:23.180 –> 0:51:25.610 Nathan Lasnoski We can bring that are possible to your executives. 0:51:25.740 –> 0:51:33.390 Nathan Lasnoski We can help you with doing the envisioning sessions and ultimately building AI enabled applications to create value for your organization. 0:51:33.710 –> 0:51:36.330 Nathan Lasnoski So go fill out the survey. 0:51:36.620 –> 0:51:39.110 Nathan Lasnoski Indicate how we can help and we would love to do it. 0:51:39.120 –> 0:51:41.320 Nathan Lasnoski We will do those first three steps on us. 0:51:41.370 –> 0:51:44.310 Nathan Lasnoski So we will do envisioning on us, we’ll do executive meetings on us. 0:51:44.320 –> 0:51:48.180 Nathan Lasnoski We’ll help you with creating and selecting the right scenarios on us. 0:51:48.190 –> 0:51:51.100 Nathan Lasnoski We want to help you gain forward movement with AI. 0:51:51.170 –> 0:51:53.140 Nathan Lasnoski That is the reason for our business existing. 0:51:53.310 –> 0:51:57.620 Nathan Lasnoski We want you to gain value in revenue, revenue, production or operational savings. 0:51:57.710 –> 0:52:2.770 Nathan Lasnoski If you are enabling M365 copilot, we had a whole session on that a couple weeks ago. 0:52:3.90 –> 0:52:11.850 Nathan Lasnoski We can help you gain forward movement on the copilot space as well, and this is a tremendous opportunity for you to get the commodity capabilities of your organization raised up a level. 0:52:12.550 –> 0:52:16.60 Nathan Lasnoski So I’m super privileged with the opportunity to be able to be with you today. 0:52:16.190 –> 0:52:24.540 Nathan Lasnoski I just loved all these conversations about AI and especially love showing you things that we’re doing and customers can make it really proud of what we’re doing. 0:52:24.620 –> 0:52:28.490 Nathan Lasnoski I would love now to take any questions that you have. 0:52:28.620 –> 0:52:33.450 Nathan Lasnoski So I’ll just kind of open up the mic and would love to dig into any of these use cases. 0:52:33.460 –> 0:52:38.630 Nathan Lasnoski Answer any questions and kind of dive into wherever you guys wanna ask away. 0:52:38.640 –> 0:52:41.450 Nathan Lasnoski So we got about 8 minutes for that. 0:52:41.460 –> 0:52:48.810 Nathan Lasnoski So and I’ll go ahead, guys and gals put out your put out your questions or you can either go off, Mike or you can put it into the chat. 0:52:48.820 –> 0:52:49.740 Nathan Lasnoski Either one of those is fine. 0:53:19.10 –> 0:53:21.560 Nathan Lasnoski Wow, I guess I guess I hit all the topics. 0:53:21.890 –> 0:53:26.690 Nathan Lasnoski You, my awkward pause didn’t work and I didn’t get a question, so that’s totally cool. 0:53:26.810 –> 0:53:38.980 Nathan Lasnoski I would love to answer any further questions one on one offline or as a follow on, I am super just thankful for all of you coming and sticking around and watching all the different aspects of this today. 0:53:39.180 –> 0:53:42.60 Nathan Lasnoski The love to have a follow on session with you would love to do. 0:53:42.70 –> 0:53:46.300 Nathan Lasnoski Envisioning would love to talk more, and I hope you all have a really wonderful rest of the day. 0:53:46.310 –> 0:53:46.860 Nathan Lasnoski Thank you. 0:53:47.860 –> 0:53:50.450 Nathan Lasnoski No, actually I didn’t do the presentation. 0:53:51.30 –> 0:54:0.800 Nathan Lasnoski I did not produce the presentation with AI, which is probably something that I could have done, but I did not produce presentation with AI that would have been a nice touch for the future. 0:54:3.220 –> 0:54:3.550 Nathan Lasnoski Alright. 0:54:3.560 –> 0:54:3.980 Nathan Lasnoski See you guys. 0:54:3.990 –> 0:54:4.380 Nathan Lasnoski Thank you. 0:54:4.390 –> 0:54:4.940 Nathan Lasnoski Have a great day.