Discover High-Value A.I. Opportunities in 5 Steps with Jonas Wauters at AIE 2024

Hello everybody. Welcome to our keynote about how to discover high valuable AI opportunities. I am so excited that you're here today.

Don't know where you're dialing in from if it's morning, afternoon or evening, but welcome. My name is Jonos and I'll guide you to this presentation today. We'll try to make it interactive as well, so feel free to drop some things in the chat and make it a nice interactive presentation.

So who am I? I'm Jonos. I'm the co-founder of an organization called floify and we help improve the operational flow at our customers.

Myself, I have over 12 years of digitalization experience of which I've helped smaller ASEs to larger international organizations. I'm a me Microsoft recognize solution architects in the power platform, which is a low-code, no-code solution. Uh, since 2023 and I'm only the number one in Belgium and one of the 51 in the world.

com. On the right side you see like a QR code that you can use to scan. If you'd like to connect with me, I often post regular on a regular base information regarding artificial intelligence.

So feel free to follow me along and that I can share my wisdom with you. So for today, what are we going to do? We're going to dive right into it.

First we're going to start with describing what exactly is this AI that we're talking about. Next up I have some inspirational use cases that I wanna share with you and then we go right deep into it. How can you find the right AI opportunity ending with some tips and tricks?

Sounds great. Like let's go first one, how, what exactly is this artificial intelligence? Well, artificial intelligence is a collection of tools or computer systems that are capable of performing tasks that normally require human intelligence.

Basically what I'm saying is like a human, the capabilities of a human. Now a computer is able to do that artificially. Yeah, you can start digging can be pretty amazing.

Right On the left side you see some definitions that I also want to share here with you so we know what we are talking about. Artificial intelligence is the model of all definitions there and below you have different subset definitions. One of them is machine learning, machine learning.

Under that you have like deep learning and in deep learning you have large LAR large language models like chat, GPTI guess everybody knows and uses it today. You also have something that's called data science, which is a cross disciplinary thing to to perform data analysis. So artificial intelligence really saying a lot and you have different kind of subsets until you go to a large language model like chat GPT.

But actually it's nothing more than a lot of math behind the scenes. You can see here a picture on the right side of a human brain and the brain is connected with neurons and electrical signals. On the left side of this slide you can see the different mathematical representation of potentially a brain.

You can see as well here neurons and the electrical signals that can flow from left to right. So it's a lot of math behind the scene. And how does a model like this train, how does it get smart?

Get down here in the chat. Who has the answer to this question? What do they say?

Data is the new what eh, anyone knows what people often say about data. Drop it down here in the chat. Data is really the important ingredient for AI training.

The models say what you previously saw with the maps. It needs data as inputs and data as outputs to compare and based on that data it trains the neurons and the model as such. So data is really, let me look at a chat if anyone has the answer already.

Yeah, gold data is the new gold they say. Some people also say oil data is the new oil, but I'm more a fan of these goals. Now data like chat, GPT, it's a big large language model.

Everybody knows it, everybody uses it as well, I'm quite sure and it's trained on a lot of data, data from the internet, Wikipedia, social media, news articles, papers, forums, a lot of data and we don't even know which data sources actually and this is also causing some concerns and some copyright lawsuits. You see a lot of bigger organizations like the New York Times or even Elon Musk suing open AI and Microsoft for these potential copyright claims. And yeah, don't worry, I always say to our customers because Microsoft and open ai, they're like, we'll we got you covered.

If there's any copyright lawsuit that pops up, no worries. We'll cover the costs of that lawsuit. That's how confident they are that it's not going to be a problem.

But in parallel you see those organizations like Microsoft OpenAI but also others like Google getting new deals, deals with proper data, providing organizations like Reddit for people that know Reddit, it's a popular forum online where people have a lot of conversations. So it's really a very good data source for companies to train upon. And you can see here an organization like Reddit has acquired a deal that says 60 million per year just for Google alone.

So that sounds pretty gold to me. Now I wanna as well emphasize the transformation capabilities of artificial intelligence on our society. Imagine yourself that you're back 100 years ago and you were working in the field and there was nothing existing like electricity.

It was, it was new, some suddenly there was like a light bulb and there was like ah, now we can see in the dark. Like you could imagine at that moment like, ah, this is like quite amazing. We can see in the dark right now.

But this was just a tiny fraction. Could you imagine back then that we are now talking live during a webinar online worldwide? Me sitting in an air conditioned room with lightning everywhere, with microphones, with cameras.

Could you imagine that? No, we couldn't. It's the same thing for artificial intelligence.

We cannot state the transformation capability that it'll have. We can only be certain that it's going to be huge. So we we expect a similar transformational power like we had with electricity.

So this can be pretty wild. And we are now at a turning point where the difficulty of AI product projects has been going down year after year after year and really had a pivotal moment with the release of chat GPT before we had like big organizations like a Facebook or a Google which had a lot of money to invest in ai for example. Those algorithms to keep you scrolling.

Those are all based on artificial intelligence and artificial intelligence is nothing new. It exists since the 50. It's just year after year after year that the cost is going down and the value is also going up.

You can see with generative AI like chat GPT, the amazing capabilities that you can now suddenly do with artificial intelligence. So we're now at a turning point. It's now the time to embrace artificial intelligence because of the fact that it's so much less difficult and so much more valuable.

Alright, I go, this gives you a bit of an introduction on what is artificial intelligence but also gets you some inspiration like okay, yeah we need to get going with this and what are our customers doing with artificial intelligence? Well a lot of different things and I'm just going to give you a couple of inspirational cases here now with you today. One of them is image generation.

Imagine before you had to hire a graphic designer or even a photographer if you wanna do some product placements and searching images online and making sure you're not copywriting anything was kind of a hassle, right? Well now as an example here, imagine you're selling sofas. You no longer need to have a Photoshop person Photoshop in your sofa in some environment or even worse, getting your sofa placed in a specific environment.

Imagine the costs. Now there are tools using artificial intelligence that actually even on demand lets you position your sofa in the environment that you want and imagine your customer having the ability to take a picture of their living room and seeing your beautiful sofa positioned right there would be great, right? Other examples from image generation part, very popularly of tools like Dell E but also mid journey.

Those are very popular tools to create images like the images you see on the bottom right here. They're all generated by ai. These persons don't exist in real life unless the pope, yeah, he exists but he never wore that jacket.

Or same for Donald Trump here. He never was involved in this. So you can see it is getting pretty wild as of the capabilities of an AI generated images and also potentially with fake news.

So watch out what you see on the internet nowadays. Another popular example is a chatbot and we have different kind of chatbots that are going to be uh, that are useful at our customers. One of them is a lead capture chatbot.

Very popular scenario in this case for an insurance company where it allows you to avoid a lot of back and forth emailing for creating like a quote for an insurance. This was an unpleasant user experience. Now with the chatbot, the potential customers can really have a conversation and make sure that the right questions are asked in a natural way so that you can deliver a quote much faster and you have a lot of those chatbots existing nowadays and a typical pattern is like chatting with your own data.

So a chatbot on its own is nice and fun, but it's like it really becomes transformative if it can chat with the data that has with your company data. Imagine you get a lot of questions from an HR department like, okay, what is my remaining holidays or what is this insurance policy? Am I covered for x?

Y is that you can imagine a lot of questions or even from customer support, my printer is not working and looking at a certain knowledge base article, imagine the chatbot is able to properly share, search intelligently search and provide you with an answer. It would save you so much emails, so much hassle and it could really optimize your business a lot. It doesn't stop there.

You also have video and ai, one of the popular and amazing tools that I like. It's captions. It really allows you to create your own videos, your own montage with different and typical captions created on the video.

Very powerful tool. Data extraction. We also see a lot of use cases there where you go for unstructured into structured data, the vision or OCR capabilities of the model today.

It's really mind blowing. You can try it out today. Take chat GPT, take a picture and ask GPT to describe the picture or describe what it sees.

You'll be amazed of the things that it sees that you didn't even notice. Try it out. Other examples, creating of your own songs.

You can create your own songs nowadays, but also your avatar in the middle there. That's a person that's my co-founder Dieter. He's actually being generated as an avatar.

He is not real life making his voice or recording there. No, it's an avatar that we actually write some text to and I can make Dieter say whatever we want on the right. Even video creation, like open AI's Sora model still to be released but it's upcoming.

Imagine just providing some text input and getting a video as a result. Bottom left there you have the virtual employee or agent experience. These are your custom gpt, also very powerful and the robots are coming my friends, this is a picture below here of the founder of uh, Nvidia or the CEO of Nvidia where he was presenting on stage and he was, yeah, with a lot of different robots on stage.

And you can see this is the Figma one. The capabilities of those robots is really going at rapid speed. The development is going very, very fast and the robots are really coming on the right side.

You see a Figma which has been connected with a GPT model chat, GPT, and the conversation is having a natural conversation with the robot and this robot in this case is cleaning up some garbage but also doing the dishes. Let me know down in the chat, who wants a robot in his house doing the dishes? Well I would be one of those but unfortunately or fortunately, depends how you see it.

Those robots are not here yet. It's not that we can buy them already for running around in our living room here or doing our groceries or our dishes. They're not there yet.

It just takes a bit more time. But these robots are already here and it's called hyper automation. And best of all those automations, they are like don't taking any holidays.

They don't require to be micromanaged. They deliver fast results 24 7 and they're not looking for a new job on LinkedIn. It's an automation that runs in the background as a robots.

And examples of these hyper automation is, let's say this one, the automation of social media presence and posting. Imagine you have like a WordPress website where you post like a new post and you want to summarize this post using chat GPT and depending then on the different social media channel, you wanna post it on Twitter with a more shorter version of it. You wanna post it on Facebook maybe with an image and you also wanna post it on LinkedIn.

Well, you can automate that with the hyper automation capabilities if this then that. Now how to find the right AI opportunity. That's why we're all here, right?

So we'll explain you. We're using a simple five step framework and we'll take you on a journey. First it starts by mapping out your organizational structure.

Secondly, zooming into pain points. Thirdly, looking at gain areas. Fourthly, looking at the AI capabilities and finally looking at innovation sprints.

Let's go one by one first. This is about mapping your organizational structure. If you have a smaller organization, you can also talk about roles that you have in your organization.

For larger organizations, you can even have advisory boards, you can have different management or CEO level, uh, different type of structure. It's important to start with mapping them out. This is an example of a typical organization that we see which has final finance, sales, hr, marketing, accounting.

So start by mapping them out. Next, when you have them mapped out, you need to define, okay, which of those departments I would like to actually take a look at for automation or artificial intelligence opportunities. So choose the departments that you want to take in your scope once you have defined that plan in a workshop and we can help you and facilitate you with the full process.

I'll share a bit more information at the end. But the important part is that you then planning that internal workshop with your employees from the relevant departments. And then you go together in like a workshop highlighting pain points.

That's where you start. Problems, challenges, obstacles, what is the current business facing? And you can have people from HR in the room, but also the other people from the other departments.

They can also have ideas. Maybe they also have pain points working with HR even though they are not hr. Feel free to drop here in the chat.

What are, what is the potential pain point for you in the HR department? I wanna see some examples. Go ahead and drop some of them in the chat.

What are pain points in the HR departments today? I'll give you some examples that we see. One example is here, employees asking the same question over and over, like how many holidays I have remaining?

Does my insurance cover this? Does my insurance cover that? All of those questions are often being asked over and over again.

Can be a pain point. Second one, lots of resumes to screen and hard to perform. A pre-selection for a given role.

I imagine you have to screen a lot of resumes. It's very hard to find the right resume for a different, for a specific role. Looking forward at other uh, pain points that you have.

Feel free to drop them again in the chat and see from each other what you guys and girls are thinking. Next up is the benefits and the gain. And this is typically a more difficult exercise so it takes a bit more time and the pain areas that come across like immediately like uh, this hurts.

But what are potential opportunities? What are benefits or positive outcomes that you would like to achieve? Same here.

Looking forward to whatever you guys are thinking of. Please feel free to drop it down in the chat Again, giving you some examples. First one, what if we have a personalized onboarding experience with timely support for new employees?

Uh, that would be great. That would be a nice benefit in our organization. Second, what if we improve our efficiency by automating some of the routine tasks?

Like it would be amazing like we have a learning and development platform. What if we could automate some of those emails? Uh, this would be an amazing game.

Nice. All right, and other ones again in the chat. Then step number four.

So we mapped out the organizational structure. We talked about the pain areas, we talked about the game areas. Now we are going to look at the AI capabilities and this is where we have mapped out different capabilities that AI can offer today.

And they're ordered, they're specifically ordered. The one at the top is more easier than the one at the bottom. Let's say the last one, controlling machines and robots.

This is still pretty complicated today and complicated can mean you need quite some budget and quite some time to achieve something there. Where at the top one, finding and organizing information. Imagine chat GPT, it's really very easy and straightforward today and you go one by one through these AI capabilities and by looking at them you try to come up with solutions.

So imagine hr, you have the pain points, you have the game areas in your head. You talk about finding and organizing information. Like what if we have to resume that?

We search the resume for the relevant skills we are looking for, we expecting this could be something and we are looking at the resumes and finding the relevant skills needed. So that's find and organizing information. Next one, perform simple tasks.

What if, ah, what if we send an e modification when a new item course is added to the learning platform? Now remember that example, we might create acasa, perform simple tasks that AI capability we could leverage. And thirdly, remember when people asking the question same question over and over again, well we could use the chat and talks capability.

So creating an employee HR chat bot that chats with your company data. For example, your policies, alright, we got some solutions. We have mapped out the solutions based on AI capabilities, based on your pain areas, based on your gain areas.

Now what's next? Now it's important to determine the right initiative that you would like to pilot after you determined IT initiative. You can talk with an AI expert, that's someone potentially in-house or maybe a partner like Floify, like what we do.

Then thirdly, you deliver the proof of concept as fast and as cheap as possible because you wanna prove the concept. And lately lastly, when the proof of concept has been proven to be working, you go to production to deliver the actual value. Now how do you determine the right initiative that to pilots?

Well this is where we give you an easy matrix to use. Remember the order of the different AI capabilities and that determines the complexity or the cost. That's what you see on the horizontal axis here.

The cost on the left side, low cost on the right side, high cost. And then on the Y axis, the vertical axis, you have the benefits low high, the top left quadrant. This is really the sweet spot.

This is where we have the quick wins that initiatives that don't cost a lot but deliver us a high benefit. It's the first initiative to start with should be in this box. So let's go initiative by initiative.

The first one was the resume screener. Looking for the right skills. I'm putting this here, it's not too complicated to do so.

So cost is relatively low. But benefits, if we think about it in this example organization, we're only screening a couple of resumes per week. It's not worthy effort to automate that.

You don't need to do AI just to do ai, right? So it's a lower benefit one. So not something that has a priority.

Next one, the EMO notification. Slightly more complicated, a bit more work, a bit more effort involved, but also more more benefits because of maybe hypothetically we sent quite a lot of emails manually at this moment when something is added to the learning platform. But then the sweet spots, this is where the employee HR chat bots can come in place.

It has a huge benefit because the amount of question that you as an HR department receive is quite substantial and provides us quite some benefits. So the employee HR chatbot could be the perfect initiative to start with. Alright, that was kind of rapid fire speed.

I hope you guys are still and girls are still following me. I'm now going to end up with some extra insights, tips, tricks and pitfalls. Um, learn from our mistakes.

Don't make the same mistake that we've done in the beginning that we saw organizations doing. Take them along. First one is the not having a strategic plan in place.

Just doing something right that is a typical pitfall and you can say, well we just want to experiment at this moment and that's fine. Just write it down in your strategy, in your policy documents. We have templates, we can help you get started.

Don't worry, but at least that it's clear that the organization knows, okay, we are experimenting with AI and we're going to make a decision at some point what's we're going to do with it, but define the strategic angle to it. Second, this is what we try to solve here in this presentation, is not starting with the right AI opportunity. And by doing that you might start, start with the more complicated, more costly ones where you have to invest quite a lot and maybe return on investment was not what you hoped for.

And this might be like, well the AI thing is not something for my organization. You don't want that. You wanna start with the right initiative.

Remember the quick wins. Next one often overlooked is neglecting people and change it is very important to communicate properly and often to the people in the organization. A lot of people are quite fearful still of ai.

They say it might take my job. Why should I even collaborate on this AI initiative? That's a lot of the thinging that we see.

Now, if you as an organization, if you state that well, we're doing this to create more revenue and to increase your pay as a result, it might change people's minds. Another way of looking at it is you can say, well, the time that we get free by using ai, dear employee, you can use it. We'll start to work with a 30 hour work week as an example.

Why wouldn't you as an employee be excited to work on that? So don't neglect the people. Don't neglect the change.

Another one is risk and compliance. Today anyone with a credit card can easily go to chat GPT and open a new account or to other tools online and start spreading your data everywhere. So that's pretty dangerous and can blow up at some point, especially when it's like sensitive data.

Think about it again, define it in your strategic angle, in your policy, which tools do you allow? Which tools don't you allow? If you don't define anything, people will just go at their own and go in the wild.

Next one, value and ROY. This is a real corner store of what we do at customers. We focus, we hyper focus on the value.

If the initiative doesn't add value and you're just doing it because everybody's doing it, that's not the right attitude. You need to really think about what problem are we actually solving with ai. And lastly is starting with the technology.

Oh, we wanna do something with chat. GPT. No.

Look at digitalization opportunities. What are problems? Again, pain areas, gain areas and maybe chat.

GPT can be a solution for some of those things. But Don start with the technology in minds. Now, a good coach to us is actually the new goals.

And this is what we do at floify. We coach organizations and we help position and benefit AI in your operational efficiency. We're founded since last year and we are fully focused on AI and we want to keep things simple.

What we do and how we help you. We provide CTO as a service where we just make, you don't need to worry about anything from a technology perspective. And we make sure we put the right investments at the right time to make sure you follow the lost technologies evolution that are applicable.

Another one is strategic AI consultancy. You don't know where to start with ai, you don't know how to implement it. Policies and strategy, we can help you with consulting there.

Then as well. Implementations of automation, any automation, uh, capability, hyper automation, if there's the net workflow, digitalization challenges, we got you covered. And finally, but not least is education and training.

Don't trust our word on it. We have a lot of references. Couple of people here providing some of the feedback of working with us as an organization.

I want to end up with the AI design Sprint, which is an AI opportunity mapping. And remember the five steps we just explained. We can help you and guide you through the process.

And for this, we're using the AI design sprint, which is actually coming originating from Google's design Sprints, but a certain organization, 33 A has made an AI version of it and it works beautifully. We have cards, different sets of cards that we use to work with people from the C level people that have leadership team, uh, involved those people. We bring them together in a workshop.

It's a half a day workshop where we help people understand the potential of ai. We also make sure that you don't need any technical expertise. The cards describe different AI capabilities that are out there in a non-technical way.

You don't need any previous AI experience. And it works for small medium to large organizations. We do this remotely, like with using tools like uh, Figma, but we also do it physically offline, where we come together and we really work with post-Its, and it works quite well.

We prepare it, we have an actual workshop and at the end we also provide you with some additional deliveries. What you get is a management report where you know what is the right AI initiative to focus on. You get all the material that we used and created.

We give you some inspirational videos of all the different AI capabilities to inspire you and to get the solution and the juices flowing, ending up with concrete next steps management summary without next steps, you're not anything with that, we can help you go to the actual proof concept, base innovation sprint and create do AI value in your organization. That was a lot to cover. We made it.

I hope you find this very informative. Feel free to get in touch. I I am opening my agenda for all of you to have a free, no, no strings attached strategy call with myself.

You can scan the QR codes that's here on the screen. com and I'm looking forward to meeting you in a call, but I'm also looking forward to understand what you will be doing with ai. Again, thank you everybody for your attention.

I hope you have a lovely remaining uh, conference and see you soon. bye-bye.