Amir Haramaty: 0:03

The consensus so far indicates that 2024 is gonna be another banner year. But the hype will be gone, and it's gonna be a very practical, very sober approach when people are looking, okay, great, but how this AI can help me in my business? How can I help AI use to impact my bottom line, my ebitda?

Tom Raftery: 0:24

Good morning, good afternoon, or good evening, wherever you are in the world. This is the Sustainable Supply Chain Podcast, the number one podcast focusing on sustainability and supply chains, and I'm your host, Tom Raftery. Hi everyone. And welcome to episode 16 of the sustainable supply chain podcast. My name is Tom Raftery, and I'm excited to be here with you today. Sharing the latest insights and trends in supply chain sustainability. Before we kick off the show, I just wanted to reiterate the new messaging function that I built into the podcast. If you look in the show notes for this episode, at the top of the show notes of this episode or any episode. You'll see a link saying, send me a message. If you click that, it'll open the messaging app on your phone or your computer, and that'll allow you to send me a message directly. You can ask me questions about the podcast. You can make suggestions about the podcast. You can just say how much you love the podcast. Anything at all, feel free to do it. Don't forget to leave your name when you send the message because the messaging app that I'm using cuts off your phone number. So I won't see your phone number for privacy reasons, and I'll have no way of identifying you, if you don't leave your name. Okay. With that, today on the podcast, we're talking to aiOla. And we'll be talking about. AI in supply chains. In the coming weeks, we'll be talking to Elevate about Southeast Asia and Asia in general supply chains. We've been talking to FocalPoint about supplier management and we'll be talking to fictive. So those are all coming up in the next couple of weeks. But as I said today, We're talking to Amir from aiOla. Amir. Welcome to the podcast. Would you like to introduce yourself?

Amir Haramaty: 2:15

Thank you, Tom. It's a pleasure to be here. I like to call myself a serial problem solver. I've been in tech for many, many years trying to constantly find some massive problems come with a clever idea to solve it, build an incredible team and culture, and then deliver value to our clients measurable value. And I've been lucky enough to have. Enough I made many of mistakes don't get me wrong, and plenty of failures, but I got 15 several times I got it right. So I've been very, very fortunate to, to lead the four tech companies, to significant outcomes prior to this one, and is a great Latin phrase that I'm poorly gonna translate to English. That says that every step that I took in my life brought me to here now. And that's exactly what I do in aiOla right now.

Tom Raftery: 2:56

So for people who are unaware, Amir, what is aiOla?

Amir Haramaty: 3:02

aiOla. First of all, it's not aioli or anything else that you can eat. It's actually stands for AI operating layer. That's what the name stands for. And actually when I teasing people, and you can see that in our logo, when you look at, look at the aiOla, the most important letter is not a or i, it's actually the O in the middle. And it's not an o it's a fat intelligence loop, evergreen and ever learning. And that's really the essence about what we do here. Ai the most used and abused letters in English language at present time it's been very sexy and very trendy, and everybody talks about, you know, 2023 was the Year of Gen ai and a lot of great questions about what's gonna happen in 2024. And the consensus so far indicates that 2024 is gonna be another banner year. But the hype will be gone, and it's gonna be a very practical, very sober approach when people are looking, okay, great, but how this AI can help me in my business? How can I help AI use to impact my bottom line, my ebitda? And that's exactly what we do. So at the end of the day, what we doing in aiOla from the very beginning is the ability to connect the power of AI to the masses. So anyone, regardless of where you are in your digital transformation journey. Or it, wherever you are, you can use it tomorrow morning with taking care of the most important part in communication, which is speech. Speech by far is the most natural way, the most effective way the richest, fastest color for you can add many more adjectives about how superior speech is to any other communication means. But when it comes to processes, right now, when it comes to ai, Speech is rarely used for AI because the lack of accuracy, the poor accuracy, which comes from specific languages, specific accents, specific acoustic environment. But above all, when you deal with business or industrial processes, those are heavy on jargon of the specific industry in a specific location. And in all fairness, there's no technology off the shelf that was able to handle it. So what aiOla did, and this is really if I put a title on top of it, we were able to take any language, any accent, any acoustic environment and any specific jargon. And while everybody talks today about LLMs, large language models, et cetera, et cetera one day I was looking at my coffee. I. And I say, you know what? I don't need to boil the ocean in order to make coffee. So I don't need to go and brute force and then try to narrow down to specific, you know, I'm originally from Israel and you know, the first language I learned was Hebrew and the most biblical language. And in Hebrew we do everything in the wrong direction. We right it for right to left. So I, so I decided to look enough in the wrong, you know, kidding aside, say, Hey, what if I'm gonna start with a specific workflow, with a specific process as near as it gets, and actually I don't need large language models. All I need is that specific workflow and I'll take it. And I use that and use generative AI to generate millions of samples, synthetic samples. So that specific workflow that actually will eventually allow me with Jargonic, which is the platform we build. To take that process and build a language model specifically for that specific workflow, specifically in that jargon, specifically in the very specific acoustic environment, accent and language, and automate the process. That's what we have done, and this is a massive, massive development patents associated. But we completely, you know, there's a great article that came out by Vinod Koshla last Friday,

Tom Raftery: 6:41

Right.

Amir Haramaty: 6:41

who talks about the importance in speech for ai and talking about the fact that eventually every human machine interface will be voice first. Talks about that, you know, once an organization will find out that the process that previously took three minutes to complete can be done in 15 seconds by speech, it's gonna change everything. And eventually, and it's a tall order. That speech first, voice first actually going democratize the use of technology. Because it'll allow to anyone with any device, at any environment to do it. And we do it not just what Mr. Khoshla said, and it's very, very accurate, but actually what we're doing on top of it, we didn't stop here. We actually using the paradigm shift to unlock the opportunity to digitize previously undigitized industries and sectors. And that's the reason we are very excited about what we do here.

Tom Raftery: 7:36

We, we'll come to that in a second, but first I just wanna, be a little bit skeptical, I guess, because we're all used to voice assistants. The likes of, I, I don't wanna say their names now in case I trigger them on people's devices who are listening, but you know what I'm talking about. The ones from Apple and Amazon and Google, et cetera. The results from them are mixed at best to, to, to put it mildly. So how have you improved on that? I mean, if I were to ask aiOla, you know, the results of yesterday's, well, Saturday's six Nations rugby match, would it know that? I'm guessing not because it's not a sports ai, but you know, if I asked that to the other assistants, the chances are they would say, I'll have to send you to the web to look for that, or something similar.

Amir Haramaty: 8:32

It's a great question Tom, and thank you for asking it because you know, welcome to my world. Try to speak to one of those systems within Israeli accent. so been there, done that. But kidding aside, again actually ASR automatic speech recognition has been around the world for dozens of years. And one of the things whenever I do, whenever I try to solve a massive problem is trying to avoid reinventing the wheel. Whatever's available off the shelf, I'll take it. When I came with that concept, and I'll share with you how I got to that point, because initially my thinking was about actually sales. You know, whenever you meet a client you or any salesperson will be exposed to tremendous amount of critical human intelligence client needs, competitors pricing, DNA, et cetera. There's no argument that there's no other point in time when you're exposed to so much critical information. Then ask yourself a question How much of that is actually being captured and then you scratch your head and say, very little CRM has been around the world since 1999, commercially phenomenal success. Operationally, if I may say so, a colossal failure. I haven't found a single person say, oh my God, Tom, I'm so excited. I'm gonna log into Salesforce or any other CRM. People do it because they have to, not because they want to. They treat it as a chore, and therefore it's reflected in a quality, quantity timing. So you build this massive IT infrastructure and then a small stream of dirty water are flying through the pipes. So what have we done? So the initial concept, I had in mind say hello. I just spoke to Tom. We talk about his beautiful hat and as I'm walking right now, I know he's from Cork and he still lives there. And now he lives in Valencia and and and so on and so on in Spain. And but before I forget when I walked to the elevator in natural language, I'm telling aiOla what I just learned. aiOla would grab it, will integrate it to the CRM of choice and create this evergreen ever learning flow. And actually, Tom, that idea to off like there's no tomorrow. Two, three months later, I'm traveling February 24, 22. Almost two years ago the war in Europe broke out.

Tom Raftery: 10:34

Yep.

Amir Haramaty: 10:35

And I started to think about the implication and it's, you know, the covid in and out. Who's gonna, who, who knows what's gonna happen when China in Taiwan, who knows what's gonna happen in the Middle East Europe, et cetera, et cetera. And you say, you know what? We arguably facing one of the biggest perfect storms in modern history, geopolitically and macroeconomically. And the outcome is that it'll dictate to organizations to make a clear separation between nice to have, versus must have. Whatever is perceived to be nice to have, forget about it. I don't have time for cosmetic surgery. I have to stay alive. And I realize as much as I'm proud of the technology we develop and the solution we have, I may be considered rightly or wrongly, perception is reality for like it or not as nice to have. And you know, in psychology there's the Maslow pyramid of needs. So I created them aiOla. From aiOla's perspective, Maslow's from aiOla's perspective, which are the verticals which are immune to the, the shenanigans of the world doesn't matter what's going in the world, we have to eat food industry. Unfortunately, you have to take our medication, pharma, logistics, supply chain super critical, energy, semiconductor you can identify. So I decided to apply our entire technology to must have verticals on critical processes where we can demonstrate quantifiable RoI at scale. That was the blueprint. And I worked very closely with McKinsey for many, many years, so I'm very well connected to many of the Fortune 1000 companies. And I spoke to a Fortune 60 company, food manufacturer, and I said, when you start your day, just flip the button and start production. I said, oh, of course not. I have to complete this highly regulated pre-op inspection, and

that's 20 pages, 04: 12:06

30 in the morning. A group of inspectors have to do it. And I said production is idle for those two hours and he say yes. I say, well, what if turn the entire process speech based, hands-free, paperless. They're gonna walk and talk and you love it because say, one, it's gonna be way more efficient. Second is gonna be safer, third is gonna be smarter, four is gonna be collaborative. And now going back to your question, I believe I can use whatever automatic speech recognition off the shelf. They decided to do the pilot in Australia. They send me a recording. I put it on one of the very best off the shelf, one of the companies you mentioned previously, and I could not believe when the accuracy we got was only 52%. First I thought maybe it's the Aussie accent. Well, while it didn't help, that was not the reason. Send acoustic environment solvable. What turns out, and I'm getting to the point, more than 50% of the vocabulary was not English, was jargon. And that's when I realized we need, we don't need to talk about six nations. We don't need to talk about weather, we don't need to talk about the rain or anything else. Every process is 2000, 3000 words max. More than 50% is vocabulary. And if I can build a language model that automatically will take care of the jargon, will take care. Fresh example, last week we're speaking for one of the largest tire manufacturers in the world from South Korea. And basically what I'm pitching about the idea we're doing, the CIO gets online and he says, listen you're preaching to the choir. I'm putting words in his mouth right now. We realize that speech is the easiest way to communicate. We wanted to turn as much as we can from our processes to speech first, and we took every platform off the shelf and going, I can put numbers on what you your gut told you about accuracy and the maximum accuracy we were able to get in Korea was 60%. So I don't know if it 'cause of its Korean or accent or acoustic environment, but 60% is like zero. I cannot use it. So, first of all, thank you because one, I don't need to convince about the need. You smarter than us, you thought about it in 2018.. Second, I don't need to tell you how difficult it is. You tested every possible system. Is that correct? So let me show what a due difference. Unfortunately, I don't speak Korean. None of my team members speaks Korean. We took a recording in Korean, highly jargonized. We ran it in one of those platforms and we ran it in our own platform. And yes, we got 59% accuracy, very similar to your 60%. But when we run it in ours, not only that, we get higher accuracy, we got 93% before training, we actually realize that it's not about transcribing. It's about separating the noise from the signal. So actually we develop a technology that deals with keyword spotting. So even when I'm able to trans to transcribe what you say very, very accurately, I'm less interesting about what you add for lunch or what you add for dinner. I'm more interested about the critical data, which I need for my process. And that's really what, and when I showed was 93% say, you know, 60% is zero. 90% is a hundred. Okay. That's the difference. And therefore what we saw as a challenge you and I experience in a daily business, we view as an opportunity, and rather than boiling the ocean, we solve it specifically automatically to one flow or process at a time with accuracy, which is north of 95%.

Tom Raftery: 15:19

Okay. And is it also a learning environment? As in if it comes across a jargon term, for example, that doesn't recognize, does it kick it out to somewhere to report and say, came across this term, not sure what it is. Can you teach me this so I'll know for the next time?

Amir Haramaty: 16:10

So first of all, the beauty of it, it's, it's a learning system. Okay? So it's a, it's machine learning of machine learning. Okay. So it's constantly improving and as the jargon progresses, we learn more and more. And keep in mind, by the way sometimes we get very excited about the speech component, but we forget this is just the first step, meaning it's really from I put the four blocks. It's really from speech to data to action to impact.

Tom Raftery: 16:37

Okay.

Amir Haramaty: 16:37

so when I'm doing right now and taking care of the speech, now I can get 95 or north of 95% accuracy. Now I have better data than had before. That was previously uncaptured and unstructured data. That unique raw material all of a sudden allows me to get better insights and intelligence. That will allow me to make better decision. So therefore, this is the chain event that starts from being able to capture speech and in every step along the way, you have quantifiable RoI, because now I completed the entire pre-op inspection, which was previously was two hours when the production was idle. Now that it's speech based, hence free, paperless, when a walk and talk, the entire process is completed in one hour. Without adding a single person or a single piece of equipment for this razor thin margin industry, we just give another hour of production that's dramatic. Second, that raw material now connected and all of a sudden we say, oh, you know what? You produce a specific product as the first product of the day for the last 30 years because that's the way you conduct your business. But actually the data is streaming, but that specific product leaves a lot of residue. Takes longer to clean, that should be the last product of the day that we don't know what we don't know. So all of a sudden that raw material allow us to illuminate, allow us to see things we're never able to see before. And that's the beauty of it. So this is the beginning of it, but it's a learning system. So whenever you face a new challenge, and not only that dynamic changes, it actually, it's not just the accuracy or the efficiency which is very important, but now you can connect the dots beyond what the human eye can see or beyond with human brain can process and it's no longer just in this production line in Queensland, Australia. Now it's connected to all the different factories in Australia or Oceania. And then it connects to all the factories in Southeast Asia. And then you have, you can connect the dots and identify trends before they actually form. So that ability now takes us again, this is really, almost to the place that, you know, it's a very philosophical question. I'm not sure we're gonna step into this meeting, but that's gonna influence the future of work because everything is gonna be smarter, everything is gonna be it's all about data and what we're doing here, we're addressing data. So now the super powerful AI tools can take it to the next level. But it's a known fact that until now most of the data in the world is still not part of the game because it's uncaptured and unstructured. And we took care of that part.

Tom Raftery: 18:58

Okay. Can you give me, I mean, you, you've mentioned the, the food processor already, but can you give me a few more use cases and bear in mind that this is the Sustainable Supply Chain podcast. So a sustainability hook would be nice to have. I mean, we're all familiar with, speech to computer. We've all seen Star Trek and you know, Picard going Computer tea, Earl Gray Hot and out comes is nice, hot cup of tea. But let, let, let's try and, bring it down to something a little more concrete and, a sustainability win or two.

Amir Haramaty: 19:35

Perfect. So, you know, when we speak about sustainability, I don't like to talk about greenwash. Okay. Because at the end of the day, it's not about, you know, you know, the feathers or whatever clothes we put on top. It's really about the bottom line. And we have to do things better. Again, sustainability, it's not a, a trend, it's reality, and that's something that we constantly have to strive and see how whatever we do. How can it help us to make it more sustainable, more efficient, cleaner, greener, healthier, et cetera? And that's a major consideration. One of the things, you know, speaking to the pharma industry, which now we're deeply involved with, I remember talking to one of the largest pharma companies in the world, and he quiz me and he said, do you know what we do over here? Said, oh, of course you research and develop and produce. Say no. We chop woods. We cut more trees than any other industries because the paper process in the pharma industry, is far superior to anything. So we have some unique situation where we have a drug which is highly needed in the field, fully produced and not released because the paperwork was not completed. Okay, the shelves are now empty people are dying. You need it. But until all the paperwork, you know, with highly regulated. All of a sudden try to imagine you're taking all this paper, which goes nowhere by the way. So they generate this paper and paper and paper and goes nowhere. So from a hundred percent paper, which is very non-sustainable and it's a and turn all of this to green knowledge. Okay? So all of that, first of all is captured much faster. Digitized connected the dots. And not only that, because this is an industry that has to be highly regulated. It's the easiest access to slice and dice it from every possible angle and audit it from any way. It's not going back to the old files. So starting for small things like that. But taking to the next level because connecting sustainability to logistics. I'm dealing right now with one of the largest beverage manufacturers in the world and they have forklifts in the warehouse, operations are running 24 7, and there's a list of challenges, whether they're facing, they have to stop, fill up a paper form, and I have a measurement that I got from them, how long it takes them sitting idle and waiting for resolution versus now we help them to install a vehicle mounted tablet. Are talking to the vehicle, mountain tablet, and everything is solved in real time. So the efficiency, you know, in every industry you have those three letters, or four letters, KPI, in their specific industry, it's called PPLH, which is Pallets Per Labor Hour, and all of a sudden it jumped by 40% because of that. So this is just one example on that one area, but we have multiple, so, so from shortening production time, turning paper to green knowledge being able to make a lot of smarter decisions. So the utilization of spare parts, because now you can predict better. Okay? You don't need to send, you know, generators on other side of the world. So the key there, really, the four key words is about being more efficient. Being safer because actually you can spend more time looking at equipment versus looking at the clipboard. Third is the key component to working smarter because of the raw material we just collected, and fourth working collaboratively. So now it's no longer just, you know, narrow tunnel vision, just this production. Now it can look in a regional, global view, plan better, be way more efficient. All of that connected directly to quantifiable, sustain sustainability gains.

Tom Raftery: 23:04

Okay, good, good, good. Talk to me a little bit as well about the, the process because it's one thing for me as a forklift operator, for example, to say something to a tablet on my forklift, but what happens to that text once I've said it to the tablet? I mean, it's, it's an audio file, a WAV file or an MP three or something. What does the AI do to it? I mean, does it just convert it to text and if so, what happens to the text once it's in text format? So I'm guessing it be, it goes into some kind of, it's structured somehow. And then does that kick off actions, you know, walk me through that. What's the process there?

Amir Haramaty: 23:51

It's a great question. I'll take it one step even further because at the end of the day I can speak for myself. I've been a disruptor my entire career. I enjoy very much the disrupt. But Tom took me maybe too long to realize that most people in the world don't like to be disrupted. Okay. And that's when I had to eat a little bit of a humble pie. And listen more. Okay? And one of the key things I learned is, especially, you know, when I work with one of the largest management consulting companies that come with digital, I'm gonna hear those two words one more time. Digital transformation, I'm gonna get allergic reaction over my body, okay? And it's not just me, many organization. And the key there is to respect, meaning if you have a process, if you have a very successful operation for the last 30 or 50 or a hundred years, you must do something right, number one, and I'm never gonna learn what you already forgot about your business. So respect it. So I'm not here to move the cheese. I'm not here to disrupt what you do. I'm here to respect what you do, but I'm here to enhance and augment the existing processes. And that's a key philosophical approach here. This is a non-threatening approach. I'm here to help you. And it's a what's not to like if I'm gonna help you to be more efficient, safer, smarter, collaborative by doing the same thing. So what we have done, we respect existing processes, and if we have a paper process right now that you do on a fill of form, we will be automatically be able to digitize that form. At any Android or iOS device, and I'll give you a very great example. One of the largest logistic companies in the world has more than 100,000 vehicles Every morning starts with a process. It's a 40 point checklist. That checklist has even a very ambitious title. It's called a 60 Seconds Quick Look Audit. So two things I asked them. One where this data going to? Nowhere. Okay. File cabinet. Second, I know it sounds stupid, but how long does it take you to do the So-called 60 Seconds Quick Look Audit and start laughing. And they say 15 to 20 minutes. Tom. Three weeks later they got a URL and address that they click on the Android device or iOS device that opens up the very same form they've been used to fill the paper. I became the pen. Okay. And now all I have to do, and actually to complete the entire process that took 15 to 20 minutes on, 52 seconds on average. And now the information is flying from a hundred thousand vehicles in the right time. And you see how the tires behave in Miami versus how they behave in North Dakota. So at the end of the day is we stick to the process they're very familiar with, we digitize it, we automate it, the information will fly. I was talking to the American Healthcare system, as you're probably aware is broken. And with aging society right now, home healthcare become a major, major issue. And I've been tasked by one of the giants of our time in the industry a very large company, very well known figure which healthcare is very dear to his life. And when a shorting what we are doing few months ago, he said, can you take a home healthcare practitioner use case, a nurse that drive to you, you know, to provide oxygen, to provide dialysis provide. And it's a highly strict process. Highly regulated. We were able to do the entire process in the same format. In the same form, but everything by speech, from taking vitals to what is the catalog number of the oxygen tank, adjust exchange to how many hours I drove, how much I pay for the tolls, and what's the observation of the patient behavior. Think about it. It's way beyond, and what I like, what I really like about what we're doing here, I had a conversation last week with one of the leading VCs, Samir, in one end you broke the barrier, the speech barrier that nobody was able to do to take it to the accuracy that we didn't experience before. Second, it's extremely simple. You can use it on any Android or iOS device. Third, it's wide open horizontally, which means it's not built for a specific vertical or specific process. So that triangle or bleeding edge technology that applied in simplest form on any device by any user on any vertical. And we getting to the point that we're starting that the first client, it took us four months. To build a language model.

Tom Raftery: 28:11

Right.

Amir Haramaty: 28:12

The second clients we did in two months, now we're doing that in days. By June it's gonna take hours. But not only that, by June we will be able to empower our clients to do it on their own. And later on this year, we're actually getting to the point that's gonna be completely self-served. So you go online and you click language form, language recording. Attached language is to get it great. Click submit 24 to 36 hours you have an automatic language model that build automatically for your process, for your workflow, and that's really what gets us excited, because its truly subjectively transformational.

Tom Raftery: 28:55

Okay. Yeah, I can see that. I mean, it's not just that you've created a new user interface, it's that you're also capturing data that was never captured before and allowing people to get new insights based on that. Can you also kick off actions?

Amir Haramaty: 29:11

Yes, absolutely. So I'll give you an example. We work right now with one of the largest cruise liners in the world. Think about it. A cruise liner is a mobile. It's a modern Tower of Babylon. Okay? They got more than 100 nationalities working on a ship. I remember one of the science fiction movie, there was a tool called Babel bubble. Yes, Babel fish. You remember that was

Tom Raftery: 29:32

Hitchhikers Guide to the Galaxy.

Amir Haramaty: 29:33

Galaxy. Exactly right. So think about it. You know, what a vision, huh? Many, many years ago when that book was written, and now we are actually doing it. Okay? So we flattening the language barrier. And the key there is, I'll give you an example. One of the safety checks they're doing right now with aiOla is the equipment. So like checking all the life jackets. And let's assume for argument saying that 24 jackets are missing. It's not enough to block the voyage and cancel the trip, but this is something that has to be fixed. So basically you build the logic around it that you send automatic alert to everybody who needs to be informed. We 24 jackets short. You have two weeks to fix it. And after a week it sends alerts, it haven't been fixed, all the right people. This is not gonna be fixed by the 30th the month you will not be able to get to, to pull your anchor and go to the sea again. So this is a simple, logical component, but you can do way more than that. So basically we have a a client right now in the CPG retail consumer packaged goods and they have to measure temperature of the actual, the food goods inside the refrigerator, you know. If the meat goes bad, if the temperature is not right, has huge, significant ramification. So if the temperature drops below a certain degree, automatic alert okay. To everybody who needs to do it. So I love that part because once you have the data, and the only way to have the data is to get structured data, the only way to get structured data is to solve the speech problem. So that's the, the, the chain reaction. But once you have the data. You can do a lot of beautiful things, including putting whatever logic from alerts to action to measurable impact that directly connected to your ebitda.

Tom Raftery: 31:12

Okay, and can I also pipe that data into my ERP application or my, whatever else it is?

Amir Haramaty: 31:19

Beautiful, because that's one of the things, we are not isolated. We don't live in an island. And one of the things we have to respect is many organizations have many different CRMs, so ERP or data management system or platforms and actually can go both ways. I was talking to one of the leading IOT technology providers in the world, and the chief strategy officer gave me a great quote that I asked her permission to use, and I say, she said, in the IoT revolution one voice was left out, the human voice. And actually, Amir, you're turning every single person on the floor to a super smart IOT sensor. Okay, that's took it this way. But going back to what you just asked Tom. Yes. So two things. One, everything we do is simply interfaced into the backend system. When I work with that cruise liner and I showing demos with other clients and say, Hey, I'm IT, can I become your client? I said, what do you mean? He said, can we open up a ticket, an IT ticket by speech? Say, sure. Can we do it by French? Sure. Can you integrate to ServiceNow because that's the backend system. Sure. A week, a week later, they're now opening up ticket by speech integrated to ServiceNow Now, now if I flip it around, think about if I'm ServiceNow, Oracle, or BMC or whatever, and because at the end of the day, what's flying to their veins data. It doesn't matter how great the data management systems are, they need the data in order to do it. If they can now create an OEM feature, which enables their clients to capture the data more accurately, faster, richer, and deeper, it's win win-win across the board. So yes, we can interface to whatever you have, you can edit as a feature to the data management system. So they can acquire, they can edit as a feature as an OEM to IOT, but everything has to play in harmony and connect to whatever the clients have in place. To a simple API.

Tom Raftery: 33:10

Okay. Okay. And where to next? I mean, what, what are your plans for, you know, 2, 3, 4, 5 years down the road from now?

Amir Haramaty: 33:19

So one of the things, you know, as you can tell by now, we are not that smart. Meaning we know about our technology, but I don't understand the verticals I'm learning as we move. So when I presented for the first time to the supermarket chain, honestly. I didn't see where's the fit for our technology, because I don't know enough about that industry. But the time I finished a general session talking about the art of the possible and ai, their CIO flipped this tablet and showed me 10 potential use cases where this can be used. So it started by us going fishing, if you will, because hey, I have the technology where, but after having two clients in logistics, two clients in transportation, two clients in CPG, I can already build a library of a no brainer use cases because I'm always looking for three things where I can bring, the greatest impact at the shortest periods of time with a minimum obstacle possible. Because we all know success, breed success, and you need to overcome a psychological barrier. So the point here where I'm seeing myself is going from being a generalist to trying to learn about the world and use cases. Now, being able to frame it and say, okay, if you come to a specific industry, here's the library of 10 use cases. You're gonna impact your business, choose the one that you need, take it to the next level. Because so far, most, if not of our clients are Fortune 50, fortune 100, fortune 500. But it's a true opportunity to democratize, which meaning tomorrow morning, I can hit mid-size and then small size in any geography, in any language. And actually, you know, we talk about the revolution that AI brings to the table, but AI and ML are amazing tools but just tools. Unless we have the data that can be always limited by who can use it, for what to use it, and how impact you can generate, we took care. We taking care of that last mile, and that's really what excite us. So the next few steps for us is to frame it to create a simple way to, I don't need to learn. The system automatically will serve with a library of use cases where you can quantify and measure ROI and then offer it to all. Across sizes, across languages, across cultures. And I think again Tom, I think it goes way beyond business. You know, it's really going to flatten the way we communicate. It's gonna flatten and democratize. And when I talk about, I get very, very passionate because it's happening. And the time that took us before, you know, any revolution took many, many years now because of the compute power, because of unit economics, because a lot of the advancement that made, we can do things way faster.

Tom Raftery: 35:55

Yeah.

Amir Haramaty: 35:56

more than anything else, measure it, measure it, measure it, quantify it. And you know, there's so many boardrooms right now when the board comes to management and say, Hey, what are we doing with AI? And the ability to turn around and say, actually we use AI to increase our sales, or we're using AI to decrease our cost, or we're doing AI to increase our safety, and work smarter. I love it because that's taken it from the, from the hype AI to a very practical, pragmatic AI, and that's what we're doing.

Tom Raftery: 36:25

Sure. Sure. Sure. Amir, we're coming towards the end of the podcast now. Is there any question I did not ask that you wish I did, or any aspect of this we haven't touched on that you think it's important for people to think about?

Amir Haramaty: 36:38

So first of all, you did a very good job and I enjoy very much the session. Thank you for, for very, very enjoyable session. But I think more than anything else, I think you don't know what you don't know, which means especially we've been used to doing things a certain ways. And one of the things that I'm a very curious person. I'm an info maniac. You know, I never learn enough. I have so much useless information back here. But all of a sudden we have the ability to challenge ourselves and be open-minded. So the key there is really the, to build that confidence that regardless of how mature you are and how big or small your organization, actually now with this technology and then capturing the data, you can shift the way you conduct business. And I think this is a huge opportunity. It's, it's a massive opportunity, way bigger than aiOla, way bigger than just our technology to change everything we do about our lives. And I think this is really the, the. It's a truly, truly transformational experience. And very rarely it happens when the topic is top of mind. You know, look on VC's investment right now globally, they're all going like this with one exception. Anything AI goes like this and, and so it's, the timing is the technology, but all of that connected to quantifiable impact. And if you can demonstrate that then that's easy, because at the end of the day, whatever we do. You know, when I go to a client and clients say, oh, this is incredible how much it costs. That's the first question. And my answer is, I don't care. Of course I care, but I'm not chasing price. I'm chasing value. And if there's no value to generate, trust me, I don't want your money. I don't want to waste my time. But if I'm going to generate value, I'm gonna take my fair share. I say, fair enough. So first of all, I love that because it's not vendorship, it's partnership. It's not me versus you. It's you and me together chasing value. And I love that part.

Tom Raftery: 38:30

Nice. Nice. Great. Amir, if people would like to know more about yourself or any of the things we discussed on the podcast today, where would you have me direct them?

Amir Haramaty: 38:40

So simple enough. Again, one of the things we definitely love to share more. There's a lot to be shared beyond what we just had time to, to expose during this wonderful podcast. So aiola.com is our website. You can find me over there, you can find me on LinkedIn, and I'm always doing my best. You know, when, when you live in the kibbutz, in a community in Israel, you can be very smart or very stupid. That's okay. You can be good looking and ugly, no problem whatsoever. But if somebody will say that you either lazy or selfish, you're doomed. So the two areas that I'm always making sure that nobody consider me being selfish or lazy. So therefore when anyone that asks me for more information and they politely ask me for help, trust me. I made so many mistakes. If I can avoid you stepping with some of the landmines I did, I'll be blessed. And so short answer is happy to share more, happy to address people questions. And it's very easy to find me on our website, on LinkedIn, and I'm at their disposal.

Tom Raftery: 39:43

Fantastic. Amir, that's been really interesting. Thanks a million for coming on the podcast today.

Amir Haramaty: 39:47

Pleasure, thank you Tom.

Tom Raftery: 39:50

Okay. Thank you all for tuning into this episode of the Sustainable Supply Chain Podcast with me, Tom Raftery. Each week, thousands of supply chain professionals listen to this show. If you or your organization want to connect with this dedicated audience, consider becoming a sponsor. You can opt for exclusive episode branding where you choose the guests or a personalized 30 second ad roll. It's a unique opportunity to reach industry experts and influencers. For more details, hit me up on Twitter or LinkedIn, or drop me an email to tomraftery at outlook. com. Together, let's shape the future of sustainable supply chains. Thanks. Catch you all next time.