Omid Bakhshandeh - Co-Founder of Verneek: Evaluating AI Companies, Challenges in the AI Market, and Building a Network

Omid Bakhshandeh  00:00

The AI pace, something people should understand that this is just the beginning of 10x other innovation that is coming, and multiple new libraries or softwares that is going to get introduced. You want it to have a team or buy from a company that you know that they will innovate.

Callan Harrington  00:18

Your you're listening to That Worked, a show that breaks down the careers of top founders and executives and pulls out those key items that led to their success. I'm your host, Callan Harrington, founder of Flashgrowth and I couldn't be more excited that you're here. Welcome back, everyone to a another episode of That Worked. This week I'm joined by Omid Bakhshandeh. Omid is the co founder of Verneek. Verneek's mission is to build the most helpful AI, enabling anyone to interact with any environment using intelligent conversational interfaces with deep domain expertise. Omid has two decades of experience in the research and development of AI systems, having founded and successfully led various AI teams. He received his PhD in AI with a focus on natural language understanding and dialogue systems. And Omid is a published author in multiple major AI publications. I learned a ton in this conversation. It's not every day I get to talk with someone that has a PhD in AI, so we hit on many of the major questions that surround AI. We talked about how you evaluate an AI company, where AI is currently stronger performing tasks than humans, and on the flip side of that, where are humans stronger at performing tasks than AI, and one of the big ones that we talked about is where AI is not replacing humans. I thought Omid's answers to all of these were super interesting, and it gave me a much better viewpoint on AI in general. But the part of the conversation that I loved the most was hearing about Omid's journey. Omid emigrated from Iran to the US, without any network of people to tap into, and now he has a PhD in AI and has founded an AI company. It's hard not to be inspired when you hear his story. And we jump into it right out of the gate. So with that, I'm going to stop talking, and let's get to the show. Omid, welcome to the show.

Omid Bakhshandeh  02:45

Thank you for having me.

Callan Harrington  02:46

Yeah, I'm excited to have you here. And one of the areas that I wanted to kick off, you know, we talked a little bit about this, and I'm really excited to hear this, but tell us a little bit about how did you get into computers, just in general?

Omid Bakhshandeh  02:59

So I grew up in Iran. It's a country that, at the time that I was growing up, it was not necessarily the high tech company with like everyone having a computer or everyone having a cell phone or anything like that. It has changed. Now everyone has a computer, everyone has a tablet, everyone has a cell phone, but like at the time, it was not the case. Especially for me, I was growing up in a smaller city, so it was even more different. So in the middle school, I was very interested into kind of tech, generally, the idea of technology from the sci-fis and kind of movies. So you see it in TV, these things called computer, and I haven't seen a computer before in person. Literally, I have seen maybe some bank clerk's back office, they're working with something looks like a type machine, but I haven't seen one myself. So my parents like send me to a computer class. It's basically introduction to computers, right. So I'm super excited to see the first computer. The instructor had one computer locked up, literally in a cage, with a fourteen inch monitor. LG, remember. I don't know if you had any experience with Windows. So it goes to a screen saver after like five minutes. And the screen saver is the pipes. I don't know if you remember this or not. But Windows had a screen saver called pipes. And these pipes are just 3D rendering of these pipes going all over the place. And I kid you not, even now thinking about it, I giggle. But I started laughing. Laughing out loud in the class. I was super tiny. Everyone was way older than me. And I couldn't stop because, the idea of this thing all of a sudden, without anyone doing anything, just changes the state to this animation. For me, was just another level of realization. And the guy would move his mouse, and then it goes away, and it comes back. That interaction to me was fascinating. So that was the first time I literally saw a computer and to me felt the most amazing, unbelievable kind of piece of machine that one can see.

Callan Harrington  05:18

I think that's so interesting. For me, it was more, and I'm trying to think of the first computer, but it was that thing that sat in the corner, and it was you play video games on this thing. I don't know, what else do you do on it. Or there's the multiplication tables game that I'm thinking about, and how fast you can get all of that right. Which, what I think that's interesting about that is how kind of different that was in that there was one computer. So I guess, to that point, for you, when you're sitting there, was it more along the lines of: this is really cool, this screensaver is cool, I've never seen anything like this? Or was it along the lines of I've heard so much about these things? And is it just these pipes? And why do I have to keep stopping this every five seconds to just go back to these pipes? What was that for you?

Omid Bakhshandeh  06:08

So I knew that this is not a TV. I knew that this is not a scheduled program. I have seen cartoons before, right? I have seen animations before. So I knew that this is not just a simple animation, there is an input to it, right? It's a function of the mouse and keyboard and then the output is coming out in the monitor. So to me, I had this realization that this thing's happening in a different way that was happening on a TV. So I knew the background. I had like that background. But my expectation was not, oh, it can just, that advanced, even can generate these things by itself. And that was super interesting. So, oh, if that's what you can do with it, oh, I can build something like that as well. All of a sudden, opens huge possibilities in my mind about like, what it can do, all the variants that, okay, you can do this kind of thing, have programs, and then these programs do something. Then you see it in person for the first time with that expectation. It just changes the whole thing. It's like, I don't know, the first time you had an iPhone, right? Or you saw- a better one- Steve Jobs giving a demo of Firefox. It was just, oh my god, he's touching it with his hand. It's moving. The thing in your hand is moving, the thing is moving. And everyone in that room felt something magical. And that exactly because happened to me, with no Steve Jobs in the room. That was the whole experience I had. Kind of the same understanding of like, this thing is more than just a TV.

Callan Harrington  07:43

Yeah, I would tell you honestly, I know the feeling that you're talking about. And for me, it was the first time I test drove a Tesla. That was probably the most kind of blown away that I had ever been, where I just thought of okay, futures here. The technology is already real. I can't believe that. I would actually probably say, I would say potentially the first time that I tried ChatGPT, as well. I think the first time that I asked a difficult question to ChatGPT, I was legitimately sitting back. That had that first feeling of driving a Tesla like, whoa, this is way further ahead than I realized.

Omid Bakhshandeh  08:21

Yeah.

Callan Harrington  08:21

Which I know is a topic that I'm excited to circle back on the AI side, but to this point that you are mentioning, right. So this was a big moment that impacted you. Did you always know that you wanted to be an entrepreneur? Or was it you were really excited about this technology? You wanted to be in this technology, you didn't really care where, or was it, I want to start a business doing this?

Omid Bakhshandeh  08:45

To be honest with you, I was more excited to be- and build a company or build something from scratch, than even getting a PhD. Getting a PhD was just something that I wanted to do, not because of the PhD itself, but because I was very curious about the technology. And that was the easiest, fastest way. Not the fastest, maybe there's a better way of spending your time than five years in school, but like, it was a very good way of kind of getting into this. It was also an easy way to immigrate to US as well. So that was a combination of couple of things that made it happen. I also had a company, even back in time, started like a small companies, building robotic staff, selling them to people, doing consulting. So I was always doing this kind of stuff. I paid for everything I have myself. So coming to here and all of that. Most of it was through all of the work that I did, the back end. So that was always top of my priority list. I loved it, I read a lot of books at the time. There were not many podcasts, so I read a lot of books around that kind of journey of Bill Gates and other people and how they got there. So it was just for me super, super interesting kind of possibility that I wanted to explore.

Callan Harrington  10:07

How did the PhD help you to be an entrepreneur?

Omid Bakhshandeh  10:11

So I'm pretty sure like people will answer this very differently. Some people say it's not going to help you at all. And they could be right. On the other hand, if I was not going to get PhD, especially in AI, this one is very different. The exposure, when it comes to deep tech, the exposure is incredibly important. Not comparing myself, but like, Sergey Brin and Larry Page, right? They were doing a PhD when they figured out PageRank. PageRank was the differentiation between- it was a technology that came out of research for them to achieve what was Google, right. So you need- they needed that exposure to even start thinking about, oh, we can rank things, that's interesting. So if I was not part of a PhD program , I would never spend enough time to think about the problem of natural language, understanding of processing, and how to solve it, or being even exposed to the ideas around that, right, to build a deep tech company. So I think it's very different if you look at, for example, even open AI and Sam Altman's site, most of the other co founders are all incredible researchers. The team is super research focused. So sometimes some verticals, they really need that like a technical expertise that comes with it to be successful. I'm not saying that there wouldn't be any companies nowadays, that they cannot be successful without basically not having a PhD or having a graduate level, basically, background, but at the time, I would say that was the only way of doing it.

Callan Harrington  11:55

So what I'm hearing you say is for a lot of companies, it doesn't matter as much, because the technology is, it's a good idea, but actually building the technology and executing that technology is not super challenging, if you have the technical skill set in order to do that. But for something like AI, and natural language processing, and something that's very, very, very technical, and also very, very new. I mean, I'd probably throw blockchain in that category as well, as a similar one, where have to have very deep engineering knowledge. The PhD helped quite a bit. Is that right?

Omid Bakhshandeh  12:29

Yeah, absolutely. And again, if you're a business person, then you can go hire a couple of great people to help you to do that is also another thing. But at the end for me, I wouldn't end up in this path, if it wasn't for PhD. That is my personal story.

Callan Harrington  12:29

Yeah, it makes a lot of sense to me, just kind of given that especially how new it is, and everything else, right. I mean, so what I'm curious about is what were the challenges? So you had the technical side down, right? What were the challenges? When you're founding Verneek, what were those things that were okay, this I didn't expect, and this was a big challenge, this I really struggled with?

Omid Bakhshandeh  13:08

Yeah, so maybe this is very unique to me, but I don't think it is. But when we started the company, we were like year, like, I don't know, six, seven in US. And we don't have any family members, or any support group or anything like that, necessarily, with us to kind of bootstrap from. There was no garage to start from, there was no friend and family around, literally did not exist, right. So that was a challenge to kind of build out a network. So to just start to kind of with the small network that used, you have to expand it enough, so you can raise some money, you can hire people, you can bring advisors, you can talk to customers, and all of that, to build a company. So that is very challenging. And I'm pretty sure it's that challenge for a lot of people, even if they have the support system and all of that, but for us to another level of difficulty to kind of overcome. So that was one. The other one was starting COVID is not easy. You cannot hire easily. You cannot talk to people easily. You have to have all this masks and stuff. Go to talk to people, even find them. You cannot fundraising. I told you the network was limited, right? You have to expand that network. There is no event to even go to expand it, or like think about all the challenges come with that. Right? So we had to kind of work around all of the COVID restriction when we're starting the company in 2021, end of 2020 to 2021. And that was not easy. That was its own puzzle to solve. So what I'm hearing is, it's the first thing was, we didn't have a network. We're in New York City, and we don't have friends and family to be able to tap in to raise this initial capital. And then on top of all of this, you did all of this during COVID. I'd love to circle back to talking a little bit about coming over here without a network and looking to raise capital, what did you do? How did you do this? What were some of the things that- what didn't work? What did work? Or if somebody's listening to this, that is an early founder, same situation doesn't have a network, what advice would you have for that person? So I think everyone has a unique situation. I think the best advice is to understand your unique situation and what is really available to you. Do you have a network, and even if this was network, how can you kind of expand and figure this out. Of course, there's a lot of cold emails and all those stuff that you have to do, but you have to brute force it. Literally, there's no way around that. You have to just go through every single person you may know, ask them to introduce you to someone that can be helpful, and then ask them to introduce you to someone that can be helpful. And then being super authentic about what you have and what you can bring to the table. And when we were raising money, AI was not AI that is now. It was not six, eight months ago that everyone talks about AI. You go out of industry that people give you money, because you have an AI company, right? So it was really not like that. People really didn't know what actually you mean by having AI company, what does it mean to have natural language understanding. They haven't seen ChatGPT to understand what it is. So you have to be a good communicator, talk about this with the users, the people that you're basically trying to raise money from, and then we had a very compelling idea. And we could just demonstrate why it is important to do something like that. We had a good background. Nasrin and I are both PhD in the same subject matter. And we both work in big companies before, startups before, and we have this level of recognition, if you may, that comes with trying to kind of do this kind of stuff. But you have to just go through and use whatever you have, and try to basically find a champion that understands and loves the idea. And the, basically, the bigger proofs and what you can bring to the table.

Callan Harrington  17:22

If you don't mind. Do you have any examples that you could share? You mentioned there's cold emailing and stuff? Are those to investors, are those to just expanding your network with other founders? What did that look like? And what were some of the examples of those emails specifically?

Omid Bakhshandeh  17:36

Yeah, I mean, from the cold emailing, and cold emailing, to be honest with you won't be, it's not the highest impact, basically, vehicle that you have. So you have to find other things. But you go, for example, to events, I mean, before COVID, we would go to some events, have some network. And then when we are raising money, or we have an idea, we just send an email, hey, by the way, we met. Keep in touch with people that you met before, even the when the network is more small. We're working on something like this, what do you think, let's talk? Hey, the person that you may know, ask them, do you know someone that you've ever heard they have done any investment in their lives? Right? Or do you know anyone that has done that before? Can you introduce me to that person? Basically you go through that, and I'm almost sure everyone that's listening can find that this with a degree of one or two to get to someone that has invested before. And they can give you an advice. If they don't give you money, they can introduce you to some other people. And then that becomes like, whatever you have to expand from. And we did the same thing. We reached out to people we knew, they introduced us to some other people, those people introduced to other people, and then just literally brute force this network and expand it as much as possible to get to- It hasn't changed, to be honest with you. It's still the same thing. We are not looking for money, people are now pitching to us to like give us money- it has changed a little bit. But now we are selling things. And then that's the same thing. We have to find someone that knows someone can buy a product from us. So we have to expand that and try to sell to people. So it's all about going through the network and asking. Don't be shy of asking of something about people, and sometimes we bring advisors, you give them advisor shares, and then they can help you. They can become your champion, right? And if you incentivize people, they will work for you. And they're like a lot of things that you should do. We didn't go that route. But that's a very successful kind of a path as well.

Callan Harrington  19:46

Did you use advisors at all? Or did you guys not use advisors?

Omid Bakhshandeh  19:50

We have an amazing advisory board. Our advisors, mostly, are business and scientific advisors, and they help us on the business side, and help us on the scientific side, and kind of introductions. But we didn't do that, we didn't  raise our money specificly with advisor like that. Having advisors that can vouch for you, that's very important. If you have someone on your advisory board that can say, hey, I've known this person for many years. We have an advisor that- let me give you an example, Manuela Veloso. She is an advisor to Verneek, a scientific advisor. She was the, and is the, chairwoman of RoboCup, the robotic competition that I was in high school. So I knew her, she teaches at Carnegie Mellon, and we had this kind of a relationship that, not necessarily like keeping up with that relationship, but she kind of knew about us. Of course, we knew about her, as she was the creator of all this RoboCup competition. So now she's an advisor to our company, which is incredible full circle, if you think about it. Like the fifteen year old circle that we closed, but having her been part of our journey now, what she has started and got us to all this tech stuff, basically. But that is something that you can always try to figure out where you can just kind of build that. But if that person, for example, say hey, I know this company, I know these people, I know them from when they were a kid, I gave them a medal when they were like this much. And now they're like all grown up. And this is an amazing kind of positive feedback for other people, right? Because they see that someone very well known is kind of vouching for you, then they say, okay, if I don't know you, but these people know you, that gives me confidence to kind of be part of this journey with you. And that can be it can start small and grow and change, of course, all the time. But that could be very helpful to kind of building a brand for yourself.

Callan Harrington  21:46

Yeah, I think I totally agree with that. And I'm a huge believer, and you talked about a little bit, but one of the most powerful questions that you can ask anybody when you're meeting with them is, do you know anybody else? Right? They may not be a good fit, but is there anybody else that you think it'd be good to have a have a conversation with? And you really tied a bow on this with the last story, in that so many people let those relationships totally drop off. And that nurturing those relationships, right? If it's somebody that you really get along with, you may share some similar viewpoints and things like that. It's incredibly powerful to keep and to hold those relationships. So I think those are excellent examples. You know, one of the things I'm curious about is, Verneek has had a lot of early success. As you mentioned, you have now it's flipped, where investors are reaching out to you. What are the challenges now, as you're growing to this next stage in the company? What are the challenges that you're facing?

Omid Bakhshandeh  22:44

It's very interesting, and I told you when we started, no one knew what AI is, and what is artificial. They couldn't even, two years ago, if you would ask on a pub- I have done that, in public, you just ask random people, what do you think AI is? Right? And then they come up with any other thing but artificial intelligence, really. And I can give you examples that I've heard right, artificial insemination, or there are lots of them. Right? So because it was not well known. It was nothing that people kind of could play with it. They had zero kind of grounding of what AI could be. Maybe they have watched a movie or something. Many people don't even watch sci fi. Right? So that was the problem at the time, no one knowing whqt AI is. Now it has completely shifted to everyone knows what AI is. And everyone thinks that they know what AI is, everyone is expert all of a sudden. So now you're kind of getting to a lot of noise in the market, that people are just building like a simple stuff without thinking about how to improve this, or what is the next step for this kind of stuff. So they are very limited understanding of the possibilities and capabilities, and then that is causing a lot of noise. And noise is always something that creates bubbles, creates overvaluation of the companies, and causes a skewing use cases towards things that don't matter. A lot of waste of money, energy, inflation in the market on every single thing. So it just creates byproducts that you don't like. So that's the challenge that right now exists out there. So it's just you have to work more to tell people hey, there are like ten of these companies that are good. The other one million, which I bet there are now at least 10,000 companies out there, I don't know. But ten of them that like know what they're doing. Right? So that kind of level of communication and understanding is now on us to just be incredibly vocal, loud about like, hey, no,  what you're thinking you're doing is wrong. I know that this is solving this one thing, but it's going to create these 100 other problems, and then kind of communicating that in a way that people understand it.

Callan Harrington  25:05

Here's a question. And it's kind of a follow up to this is, so you mentioned that, you know, there's probably ten- and I know you're saying random numbers, right. But there's probably ten companies that really get it. And there's 10,000 or so companies, and there's a new one popping up every day, right? How can the average consumer know which ones are legit and which ones are not?

Omid Bakhshandeh  25:29

So I don't know. To be honest with you, that comes down to the end users and use cases. The rule of thumb for me personally, whenever I evaluate a company, kind of, I get also a lot of different people pitch to me, they want to raise money, they asked me to kind of invest in their companies, right. And then the way that I evaluat it myself, I look into it as like, I just look at the people. It should be one of these two, you are an amazing, like AI team, right? If you have like a good AI background, and I understand some of your work, or I know you know what you're doing, basically. So I say okay, they at least technically, they know what they're doing, they can innovate somehow. Or the other way around, like they've been in the business for payment for fifteen years, and now they've built like a multi billion dollar company in payment. Now they have an idea about how AI can help payments. And now they're trying to find a couple of AI scientists and engineers and stuff like that to help them with that process. Then I say, okay, that's the second category. So I always look at the founders to kind of measure the quality. But at the same time, that's like, looking at something that is maybe pre product or very early stage of the product. There's a company that has built a product that you don't even understand what they are, who they are, but the product is doing amazing, it works really well. Then you will end up trusting that right? So okay, this product solves my problem, nine out of the ten. And therefore, that just like product speaks for itself. And that's another one, that's another way of kind of looking at it. But if you're trying to join a company, that's an AI company, make sure the founders are some level of competency. And if you want to invest in a company that is AI company, same thing. If you want to buy from a company that has AI product, and you want to buy it from it, if you really cannot find anything like that in the market, sure, go for it. But like, if there are other options out there, definitely do this research, definitely look for one that has better team behind them, because they can innovate their way, basically. As these things get- they change. The AI pace, something people should understand that this is just the beginning of 10x other innovation that is coming, and multiple new libraries or softwares that is going to get introduced. You want it to have a team or buy from a company that you know that they will innovate as the technology evolves, and you get always the latest. You don't want to switch in the middle of your subscription or whatever you're doing.

Callan Harrington  28:17

I think that's really good advice. And if I'm hearing you, it's evaluate the team. Is the team coming from a background that can innovate, right? So here's where we're at right now, in the future, this is going to look totally different. Are they going to be able to catch up to where things are going? Or are they going to be the ones leading the way in that innovation? I've got to ask this question. I'm super curious on this. And since you're an AI expert, is there an area where you feel that the human will always be important in this particular way? Or not? I'm super curious to get your opinion on that.

Omid Bakhshandeh  28:53

There are things that are very easy for a human, they tend to be incredibly easy. This is Moravec's paradox: when something is super easy, it's kind of become very hard for machines to do that. And things that are very complex for humans, they become easier for machines to do them, towards time, as the time passes, basically. And what it means is that whenever it goes to more knowledge work, basically, you will get to a point of really high quality knowledge, kind of work. That is going to get to a point of 01 access, basically, that you can find a piece of knowledge that you want in different categories and different areas very easily. But when it comes to simpler stuff, right now, if you try to follow the latest on robotics for picking up items in like a factory, is kind of laughable when you look at the videos, how behind they are compared to like, what, how amazing all these things that we have built so far right? They're like a level of something very simple. And at the same time, even in that category, you have Boston Dynamics is like doing crazy stuff, right? They can do anything and everything, but picking, for whatever reason, becomes like very hard. There are so many little things that are going to be, for a while, a human dependent, that we don't even think about them that way, you know. Like cooking, right? There are a lot of startups trying to have robots that cook. But they are like, extremely limited to extremely factory cooking machines, right? When it comes to human knowledge, from the knowledge point of view, there are like places that maybe AI is going to be as good, if not better than, a human. But the end users or the society as a whole is not going to trust AI, or it's going to be very hard for them to trust. As a society, we feel this pressure of, hey, I need to point a finger, if something goes wrong, to a human being, right. And that is the responsibility, when it comes to the responsibility of a human in the loop for some of these tasks, that I don't think we are ready for a while to kind of trust and put this in the hands of a robot to make that decision. Or a pilot's, right? Like simple example, already, we have auto pilots that are amazing. And then we put two highly paid pilots in the every cockpit, because, I mean, they can of course, help a lot of situation. But let's assume that autopilot becomes even super good, even easier than cars that is in the air, right, to do that. But are you- for example, will you trust to go to an airplane that doesn't have any pilots? Really, I want to know. What is your answer?

Callan Harrington  31:52

It's so interesting, even how you just described that. And even knowing that, I know that it's like 90% of the time, it's on autopilot. If I'm being totally honest, and I walk into a plane, there's no pilots, I'm probably going to be pretty nervous. Matter of fact, I would be super nervous, if I'm being 100% honest. I know that I would be super nervous.

Omid Bakhshandeh  32:13

Yeah.

Callan Harrington  32:14

What about you? I'm actually super curious about that. Would you be totally fine with it?

Omid Bakhshandeh  32:18

So, at the current status of technology, I'm not even fine with the cars, to be honest. I've seen too much. I've been part of this for such a long time. I'm okay for a test. Of course, I'm okay with the test. But on a daily basis, full trust, I don't know how to feel. It's crazy, but in airports there are just trains that don't have any conductors. There is no one that does that in a small electric train. And something that if this train just stops in middle of, you know, this air train stops in the middle, who is going to help us? Who is in charge, right? I know, of course, like there's someone out there, and there is no AI even. This is like just clicks of, like there's no intelligent being here. But it feels- okay, I need to have someone even to be responsible for this. I want to know someone is in the same boat, or this case, plane, or a train with me. They have skin in the game. In the car that some programmer has written the code for, he is not here. If he was here, I would- okay, I would sit here, right? And he's not here. He's sitting behind his laptop. So why should I trust it. That is, of course not productive thinking about that. And I'm not saying that cars are going- we are- of course cars are going to be incredibly safe, there are a lot of things we go through to make sure of the safety, it probably is going to be even better, and all of that. But at the end of the day, it becomes the question of, on the responsability. Whenever there's a responsibility for something, how do you want it to kind of take care of that responsibility? And who is going to be in charge? And are you okay for an AI to be responsible at the time, right? And I really don't know the answer to this. And I don't know how fast it's going to evolve in this society. But if I'm a betting kind of man, I would bet against like a lot of places to kind of completely change to AI driven with no human.

Callan Harrington  34:21

I think that's such a good point. I never would have thought- So, essentially, what I'm picking up from that is, you have a situation where the technology can probably do it, and the test will probably show that this is probably less error proof. These are my words. I'm not saying that you said this, but the test will probably say that, statistically, there are less mistakes with the AI than there is with humans running the show on this. But to get over that mental barrier, that we're putting our trust in the AI, over the trust of a human, even if that human's only there to- just like a pilot, I think that was such a good example. A pilot is just sitting there to have their hand on the wheel for ten minutes, but during that ten minutes, we want to know that a pilot's there, so we're not going to fully trust AI taking over. I think that makes so much sense. And I think, to go back to ChatGPT, anyone that's used it, it's hard not to be kind of blown away, right? Depends what you ask it, right? It can't do everything yet. But if you ask it a complex question, and you said this, that is why it kind of clicked for me, it's like, it's already pretty good. That's only going to get better. And it's gonna get more customized, more personalized, and everything else. But yeah, I think that's such an interesting point is, will humans trust it.

Omid Bakhshandeh  35:45

Yeah. I mean, it's okay to trust it in a lot of places, because the responsibilities is on yourself. So you can be a personal responsible for sometimes  say, hey, I'm okay with it. And that's okay, then you roll it out. And people are okay with it. That's the choice that they make. But like, when it comes to things that are going to affect like a group of people then, like an airplane, or a car in a street, that becomes very different question of how you wanted to present it. And even a way to get to a point that this society kind of accepted it or not.

Callan Harrington  36:20

So, I guess, here's a question on that. How are you getting people to trust what you're doing at Verneek right now?

Omid Bakhshandeh  36:28

So what we are doing, we're building a hyper personalized, hyper localized AI for different sectors and verticals. And the way that we look into this is that we-

Callan Harrington  36:39

Sorry, real quick, can you explain what that means a little bit more?

Omid Bakhshandeh  36:42

So we are building natural language interfaces to any environment, could be your shop. So tomorrow, you walk into a shop, grocery shopping, and then you see this AI or access to it. So you can scan a QR code, or basically tap something, or there's a big machine that you can go to and start asking, hey, I want to gluten-free free ice cream under ten dollars, that is also zero calories, and there's no nuts in it, right. And you want to know what it is, our system comes back, it's hyper personalized, hyper localized to data specific to that store, to give you the answer, right? So that's in this category, could be in the finance domain, if you want to know okay, how much I have spent the last ten days on Uber. You want to just control your spending, whatever it is, hyper personalized, hyper localized to your wallet, and then you can expand this on different kinds of verticals. And we build this kind of natural language interface to kind of data on environment and actions and things that you can do. So the way we have kind of envisioned this technology from day one, was in a way that we are limiting the number of places the system can kind of go wrong and kind of do things that are not intended. So we limit the actions limit the possibilities, basically in a very specific way that makes our system very trustworthy. And then we are a kind of third party, basically provider of technology. So we are kind of like we do our best to make the technology as good as possible. And no one can say that, like, it's going to be 100%, okay with anything, right? Anything, literally anything. And you go there, and you just kind of, when we talk to our customers, we're kind of presented with all the things that come through to say, hey, these are the things we are doing. But here's what we believe you should do to even make it better and more trustworthy. Because we understand the limitations, we understand where things can go wrong, where things can go right. So we try our best to kind of make sure that we are communicating those kinds of facts with our customers. And we can put that in the hands of kind of the users that way.

Callan Harrington  39:04

Gotcha. So you're almost building it together with them, right? So one, the trust is going to be built where, in your world, because it's hyper specific and personalized to them. If it's not a good answer, they're not going to keep using it, essentially.

Omid Bakhshandeh  39:18

Exactly.

Callan Harrington  39:18

And for you to add to that to the first part is okay, we're using it. What are we missing? How can we add other things to make this more personalized to you? Is that essentially correct?

Omid Bakhshandeh  39:30

Yes, exactly. Yeah, we are a company, I mentioned, we are a super technical company. We are very deep into research, and we can innovate to overcome then, basically anywhere that a system is not going to work well. Right. Try to solve that problem. So that gives confidence to people to kind of- But at the end of the day, let's go through the example I started from, right? If a company tomorrow comes to us, and it's like a health care provider, said, hey, we wanted to replace all of our doctors with an AI, can you build this? So yeah, we can. We can build that for you, right, to a point. It's going to be, these are the limitations, these are the things you can do. Then is on, kind of the conversation becomes, is this enough? To kind of build the solution you wanted to have or not? Are you sure that you wanted to use it this way? Because these are the limitation. Are you going to kind of communicate the limitation to the users or not? We try to make sure that these are all intact. And, although we believe that we can build better than anyone, at the end of the day, it becomes the question of the end use case, right? If you're in commerce, first of all, nothing- you're shopping, right? Sure you buy something, you don't buy something, right? At the end, you are the person making a decision. It's like anything else, it's searching on a search bar, or just going someone selling, some associate, selling something to you, right, to push product to you, and then you're going to buy something. So there is like not that much, basically, scrutiny around it to begin with. So that's the part where you can be more relaxed about it, say, hey, this is it. So build it out with us, let's try to make sure that everything is fine. But if we are going to somewhere that is more critical, and they're like, building a driverless car, right? Your are the command interface to the driverless car, right? Okay, we don't know about driverless cars, you know, this is what we can do, anything after that, please understand, these are the limitations. You have to make sure that you are taking care of the rest. So that is how I think it is important when you're working with someone, that they know what they're doing. They know what their expertise are. And then to kind of have that conversation with them. That's what we do all the time.

Callan Harrington  41:45

Yeah, I love that. And it gets back to, as you just mentioned, to your point of, look at the teams. Right now, especially while this stuff is early, and look at the teams behind that, and are they experts, which I thought was excellent advice. Omid, last question I have for you is, if you could have a conversation with your younger self, what would that conversation be? And what advice would you give them?

Omid Bakhshandeh  42:06

So the advice, I'm going to give them that I'm almost sure they're not going to follow, even if I tell them this, is to kind of have more self care, and make sure that you take care of yourself. Don't skip that gym class, you know, keep doing some of those stuff, try to be healthier, from day one. It's important to be in a very good state of physical and mental kind of state to begin with. And I would say like, if you get into it from very early, third grade or something, that can like kind of propagate out, and become very easy thing, and part of your daily routine without you even knowing it. This is like maybe easy advice. I think it is very important to make sure that you have a full kind of a system around, full kind of mental and physical and everything. You have access to all of that kind of capabilities.

Callan Harrington  43:10

Well, I think you kind of nailed it, right? To me, it's simple advice, but not easy to do. Where, if you build those habits early, you got a better shot. Because once you start a business, boy, does it get tough, and it's still just as important, but it gets tough. So I love that advice.

Omid Bakhshandeh  43:25

Yeah, it gets tough, and hopefully we can- It's never too late. For people who are listening, start running today, or go to gym today.

Callan Harrington  43:36

I love it. Omid, this has been a lot of fun. I know I've personally learned a ton. I appreciate you coming on.

Omid Bakhshandeh  43:41

Absolutely. Thank you so much.

Callan Harrington  43:48

I hope you enjoyed Omid and I's conversation. I know I walked away with a totally different viewpoint on AI than I had going into it. If you want to learn more about Omid, you could find him on LinkedIn in the show notes. Also, if you liked this episode, you could find me on LinkedIn to let me know. And if you really want to support the show, a review on Apple Podcast or Spotify is very much appreciated. Thanks for listening, everybody, and I'll see you next week.