EP 448: How To Build an AI-First Company

This is the Everyday AI show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business and everyday life.

It's one of these buzzwords you keep hearing, right? How do I be AI First? How can our company be AI native? As you know, let's be honest, companies have struggled a little bit over the last couple of years since the generative AI way of wave of 2022 to really integrate generative AI and large language models seamlessly into their day to day operations. And sometimes it can be easier to start over from scratch. So we're going to learn today on how to build an AI first company and talking to the founder of Nourish. And hey, great timing because we can also talk a little bit about what Nourish is, what it does and nutrition, right? It's that time of the year. We got to watch what we can eat and we can do it with AI while also learning how to build an AI first company. It's going to be a fun conversation. All right, I'm excited for this one. I hope you are too. Welcome to Everyday AI. My name is Jordan Wilson. I'm the host and this thing, it's for you. It is your daily live stream, podcast and free daily newsletter helping us all learn and leverage generative AI to grow our companies and our careers. If that sounds like you, what I just said, you're in the right place. This is your home. Another home for you is our website your everyday AI dot com. Yeah, you can go sign up for the free daily newsletter. We will be recapping our conversation today there as well as keeping you up to date with everything you need to know in the world of AI. But also we have 400 and I don't know how many I lost track, like 430 episodes. You can go watch, listen, read and learn from the smartest experts in the world on what no matter what category, go click sales, marketing, hr, whatever you need, we got it all there. All right, so if you're looking for the AI news. Yeah, generally we do that. This is technically a pre recorded show that we're debuting live. We'll still have in the newsletter. Don't worry, we got you. So make sure to go check that out. So let's get going. I am excited for today's conversation. I think it's going to be a fun one. It's, it's, it's on the, the top of everyone's mind. How can my company, you know, go from not just talking about AI to making it the first step.

Right.

Not an afterthought. All right, so help me. Welcome to the show. Here we go. We got him. Rajesh Kandaswamy, the founder of Nourish. Rajesh, thank you so much for joining the Everyday AI Show.

Thank you for having me, Jordan. I look forward to the conversation. All right.

It's going to be a good one. So, Rajesh, first, let's talk a little bit about Nourish. What is Nourish and how did it come to be?

So, Jordan, Nourish is an AI first app. Fundamentally, what Nourish does is help you track what you eat and get advice. You know, we track everything in our lives. Money, steps, energy, use. But what we put in our body, we don't really track very well. And digital technology has helped us in many things, but somehow it has not made a mark in how we can track what we eat. Do you know that now hunger is a larger issue? Obesity is a larger issue in the world than hunger. I never thought I would see the day. This includes India and China. So we believe there is a huge chance for us to improve how we track what we eat. And previous technologies haven't really made such a mark, but we believe with AI, it's a step change, can help us solve this problem, can go beyond more both in terms of user interface and in terms of capabilities. And that's what Nourish does. It's an AI first company to help you track what you eat and get instant advice and deeper insights.

Yeah. What great timing, right? Because I think this is whether everyone's shoving their face during the holidays like me or, you know, looking at their New Year's resolutions to stop shoving food in their face. It's, it's. It's that time of year, right? So. So, you know, I'm curious, you know, how do you think, you know, Nourish or, you know, even just using AI to be more mindful of what you're eating? How. How can that be especially helpful during this time of year?

So the thing is, like, you know, so one of the challenges why this problem exists, right, why we are not able to easily track what we eat is because food, our relationship with food is very complex, what we eat. And it's a. You know, there are many apps today that can help you track what you eat, but they're very cumbersome to use. Most people try them. It works for a set of people, but most people give up using technology to track what they eat. But the interesting thing about AI and Especially Gen AI, it's a step change in user interface. Imagine like for instance, I can easily tell you for breakfast today I had a slice of sourdough and a couple of scrambled eggs and a cup of coffee. Try putting it into one of these apps, it's super hard. What if you're just able to tell like how I just told you? I can just say I had a couple of scrambled eggs and I had a slice of avocado and a slice of sourdough and half a cup of coffee. So it's so easy to tell somebody and AI can do that. And also imagine taking a picture or immediately getting instant feedback based on my goals. I happen to be a half marathon runner and then I happen to be prefer certain foods and get instant advice. So the interesting thing is why AI particularly is that compared to other technologies, AI can mix multimodalities in terms of exchanges. That's one. So whether it's speech, images, everything that we can do in the real world, that is a big step change. And the second thing is, especially with nutrition, as I mentioned, nutrition is a very complex thing. You know, what we eat is a combination of chemistry, biology, culture, our individual preferences, behavioral science, all that goes in that determines how well we keep on cap of our diet. I cannot think of any technology in the world, including all the digital technologies that can look across domains and help you that way. And AI can. And similarly there are opportunities in many other places and I'm kind of building a nutrition company. But if you are in many other places, outside in medicine, culture, finance, are there things beyond what today's traditional deterministic computers can do? That's where the opportunities lies and that makes it very interesting. That's where we are, that's what we are trying to do with nutrition.

So you know, you talked about building in there and you know, it's, it's an interesting time right now to be building anything. But you know, I want to talk a little bit about even your experience of building nourish. Right. Because. Well first, and maybe this will help set the stage. Can you tell everyone first a little bit about your background at Gartner?

Yes. So for many of you, Gartner is the world's leading technology research firm and I was there for 10 years. I loved the job and was fortunate to do many things. And one of the things that it did was to Gartner makes you a fellow and was to think about the future of organizations. So I was trying to see how organization models evolve over time. And one very interesting thing Came up. Every large industrial revolution fundamentally changes the concept of what an organization is. When we were hunter gatherers and then we became farmers, the organization unit change, then to small artisans change, steam power organization model change, metals and manufacturing cars, factory and offices came. But that had not occurred with digital till a few years ago. If you look at the inside of like, you know, a large digital organization like Amazon was not too different from a Ford Motor company. You had like, you know, CEOs, marketing, sales, et cetera. So I was trying to see like, you know, why was that not the case. And at the time, every time this change happens it is for the underlying energy of the industrial revolution. So whether it is farming or steam power or metal making electricity, but that did not occur to digital. So then as I was trying to look into that, I was trying to design something. But after that Gen AI revolution came, then I realized the underlying energy of digital was not ready for the shift and that's why the models have not changed. So now I believe the GENEA is the answer for a new type of organization that's well suited for the digital age. So which I call as the AI first organization. It's not just an organization using AI, you know, in multiple products or something. If, imagine if you're given a clean slate and you're going to start a company and you don't assume that it's driven by humans, human processes. If AI is your number one resource, that's what you're going to go to first and then use humans. How would you build that? That is AI first. And I believe the best way to do that is not trying to go and make little tweaks in existing organizations. And I believe that will grow organically through startups. And I'm hoping to build that with me and my co founder. That's what we tried to do in our company, which we call it AI first. It is anything that we do, whether it is looking at our cap table, whether it is trying to come up with the marketing Personas, whether it's trying to analyze competitor products, design like you know, testing out our strategies, developing code, testing code, putting data in all the way. We start with like you know, first with AI and who we bring in, they have to start first with AI and gen AI tools. Clearly it won't work for everything. We will augment it, but we'll keep getting better over time. That's what we are trying to do.

Right. And you know, you were in a fascinating position there at Gardner, I'm sure right where it is your Job to understand, you know, worldwide business trends and how these changes in technology are ultimately shifting how we live, how we work, how we exist. Right. But you know, I'm wondering was there any one trend? And obviously, you know, I'm guessing it was, had something to do with the emergence of large language models. Right, but what was it that you saw in your position as a fellow there and in researching all of these technological trends, what was it that really pushed you to say, okay, well hey, I'm going to build this, this nourish thing and we're going to put AI at the front and center?

I think that part of that answer is personal. Part of that is technology. The part of the personal answer is I had a long, long time ago when I was in India and just when Internet was taking off. I remember walking in early in the morning and this is before the age of browser. You could send these emails to list servers and they will respond to you with output. I need to send something to universities and I'm clearly dating myself and I was fascinated that I wouldn't consider myself as a deep technology guy at all by any means. But that was fascinating that I could send these commands to Stanford University of University of Illinois, Urbana Champaign. They will send you back. Without knowing me, it is just mind boggling for me. And that feeling, I got it again only when I started chat with ChatGPT. So and I felt that I needed to do more with the new technology. Gartner was wonderful, but I want to do something on my own. And second thing is they realized with this, this future of organization fellows research as I did, the best way to try to try that was to go and try to build something on my own. Large organizations do create a lot of value in the world. They will adopt AI absolutely fine. But for my area that I was thinking about, the best way to realize that was to start own. That's the second thing. And the third thing was nutrition. I want to do something that actually creates value to people. And clearly as long as we have in my mind capitalism and as with like, you know, more and more food being sold and manufactured and obesity being a larger problem, this is a problem that exists. I have tried to do something in my own, in my own family, trying to start WhatsApp groups of people to better track what they eat, etc. That hasn't worked very well. I myself am extremely habit driven. You know, I track everything, how much I run, how many books I read, everything. But ask me how well I eat and how do I track it. Really bad. So all of these things put together led me to wanting to start this and I met a co founder, Suresh, who is very interested in it, who's deeper into technology and we share the same vision. We said let's go do it.

Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on gen AI. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft and Nvidia have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, you can partner with us too. Just like some of the biggest companies in the world do. Go to your everydayai.com partner to get in contact with our team or you can just click on the partner section of our website will help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on gen. Looking back at it now, right, so you, you started Nourish, what, just about a year and a half ago, correct? What, what were some of the maybe AI first decisions that, that you maybe made, you know, a year and a half ago or, or more. What are some of those AI first decisions that you made that maybe at the time felt a little uneasy. Right. But looking back at it now, you're like, oh, okay, this was the right thing to put artificial intelligence or, you know, large language models at the center of this process, at the center of this ideation, right? What were some of those decisions that actually maybe felt weird but paid off?

Yes, I think it's an important one to go into. So when we started this, right, just to give the other time, this is a year and a half ago. Yes, ChatGPT was interesting, but actually using it in a product wasn't very sure. So for instance, what I just said, like, you know, I had a couple of scrambled eggs with a slice of sourdough and like, you know, half an avocado with a cup of coffee, that giving something like that to an LLM and getting the response used to take 90 seconds. Right. So at the time we made this big bet that this technology, we can't predict everything in the Future. But one thing you can predict in the future in computing is computers will become cheaper, they will become faster, connectivity will be better. So based on that is we said yes, today is going to be 90 seconds, but it's worth building it for that because technology will catch up. The closest analogy I can think of is when Amazon, Jeff Bezos tried to build an E commerce Marketplace for a 28.8 modem. Again, dating myself, it was extremely slow, but they expected Internet to become connectivity to become faster and then when it arrived, so that's a fundamental bet we took. So it took like around 90 seconds or so and now the same thing takes 10 seconds. We have done some work to improve, but most of the work is done by the technology, underlying technology and the providers themselves. So we are riding on that. Right? That is one, one, one big bet, I think, which worked quite well. The other thing we said was in terms of leading on to how we build this technology, if it is AI first, AI is your first resource. Then what you need is an aptitude to work on it and the willingness to, not necessarily your skills. We started this interesting program which is to use high school interns. What we did was we created a mechanism in which a lot of high school kids want to learn computers, want to do AI and they have the right aptitude. And these are students who are some who live close to us and some who are far away. I have high school interns in India, Dubai, China, uk. And what they do is when they start with us, they could be. A lot of them are in software development. Since we are AI first, it's all generated code and some of them are in marketing, some of them do a marketing video, some of them do content. So when they come, we don't really train them, teach them everything. We tell them to go and use AI to get everything set up and they should use AI only over time to build it and learn from others. And so that's the second bet we made and that bet has paid off. We've learned, I wouldn't say it was completely smooth, but we have learned how to be an AI first company and then use humans and what the dynamic changes. That's a second key bet that we did. Does it help?

Yeah, and, and thanks, thanks for walking through that because now, now I'm, you know, my mind is racing in all kind of different directions because I think, right, one of the reasons why larger organizations are fail, failing at implementing AI and truly becoming AI first organizations. Well, number one, they're, they're, they're Too big, too slow. Right. You can't do this over a one year pilot. But I also think it's an older way of thinking, it's an older way of working if you're not AI first. Right. Because you still think, oh, you know, our value prop is, you know, we solve X for Y by doing Z. Right. That the very, you know, famous value prop approach there. But you know, does AI like, does being an AI first organization change what's possible? Do we still have to go do the old boring road map or you know, there's all these, you know, conversations like oh, AI is going to help, you know, a one person company become a billion dollar unicorn. Should we change our thinking and change what's possible? If we can properly put AI at the center.

Absolutely. In every aspect of you as a, as a worker in a company or whether you're self employed or you know, in any aspect of what you do. Starting to use AI. And these, most of these tools are available for anyone. And one of the things I see is the people who are willing to go and try with no preconceived notions, they do so much better. Because a lot of people, I mean like, you know, I've spoken to many people over the last few years on technologies and there's not unique to AI but it's a larger problem in AI because people know what human intelligence is. They trying to always compare AI and what it can do and a machine and machines intelligence errors it can do in a particular way. And I'm saying don't just go play with it. It's interestingly what happens is it's not only AI that will improve over time. We become better at working with imperfect intelligence. If you see over time in any technologies that we use, say we see something on the screen initially it used to be black and white. It's a flat screen. Clearly it's not the real world. Clearly it's not perfect like what's in front of you. But over time we adjust, our senses adjust, we can still emote thinking. It's like real people though it's on a flat screen like that how AI decisions the content generates, we will learn to appreciate, we learn to work with imperfections while the technology is improving. Another thing which you realize is coming back to the points about using interns is if people without preconceived notions but wanting to get work quickly done. There are very few people in the world who really want to get stuff done quickly and not waste time compared to high school kids, sometimes procrastination sometimes laziness is a good thing. So they are very, very open to saying, you know, why can't I make the computer do this? You know, I was born in a world that gave me computers and then they connected the wire to it. They were born in a world in which computers came with wires, they came with phones, and their relationship to technology is so different. They're very comfortable. Interestingly, our observation is that a lot of these high school kids, and thanks to their teachers, they are more open and thinking about use of technology compared to people who are already in colleges learning computer science, they feel somewhat more threatened. Unfortunately, because of feeling threatened, they seem to avoid and putting themselves at more peril than the people who aren't.

You know, I, I love that story, right, because it even resonates with me personally. Right, Like I love that you, you know, gave, gave high school interns a shot to work something that could be bigger than themselves. You know, I'm even thinking, you know, I got, I got a job to, you know, work at a daily newspaper when I was, you know, in high school still. So, you know, I, I, I, I love that you were giving opportunities to, to the future generation. I, I wanted to shift gears here a little bit and, and talk, uh, a little bit more about this kind of trend spotting. Right, because I know at Gardner, uh, you know, that's what you were trying to do, right? Like that's what big research organizations do, right. They see what trends have already happened, what's happening now, what's coming next, and how the business world can make use of that information. So, you know, if you combine, you know, your, your decade of doing that, and now you're, you know, year and a half of trying to take that information and build something with it. What trends have you seen? So what maybe worked? And, and, and maybe do you see any disconnects? At least know what the industry analysts are saying, like, oh, here's where AI is going versus you building with it day to day.

Yes. There are two, three things I would say is one, let me get, how do I think about this? Let me just share it and I'll kind of go through two or three ones. One is, in terms of friends, like I mentioned earlier, there are some inevitable truths about how things evolve. Computers becoming cheaper, faster, AI getting better connectivity getting better, more people getting computers, all of the inevitable truths. So you pick any technology. You don't need to try to think of everything technology does along these three or four vectors that will anyway occur. If that happens, how will the technology Be that's all we need to think about. That one we can be clear about. The thing we need to be open about is at any point in time we have only one past but we have multiple futures because we don't know. So being open and not necessarily tied on to one being willing to change is the second thing that is important to keep in mind now in terms of trends of one of the things which is very important in my mind which is somewhat counterintuitive to how things are projected. When people talk about Gen AI, they talk about hallucination. They talk about how it's not completely reliable. So because of which we can't use our take on this is completely different. When we started we were initially thinking we're going to build our own database and what we and say for instance a year ago we were trying to build something with gen AI but we thought we have to build our own database because for instance, if you say I ate a slice of Apple, you asked LLMs 10 times. They used to come up with few different answers. Today ask the same LLMs they come up with the same answer and the speed has improved. So now we feel like we knew it is going to get better and it is getting better. What I want to go there is when people think about hallucination, they're thinking about LLM purely as a technology. But the trend is as technologies get bigger, they become brands. It's not LLM. GPT4O is a brand, right? A chat GPT is a brand for the company with billions of dollars for anybody to use them. They have to be able to offer the right answer because in some cases generation is okay calories in a slice of Apple, who's the seventh president of the U.S. all of that you need absolute truth. And the users are not going to separate which question I want to this question I want generation. It's the job of the company. It's job of the brand. So what they will do the brand was ChatGPT. Underneath it there might be an LLM but there will be all the other scaffolding to make the thing work. So that's the trend as like you know these LLMs you don't think only as LLMs you think about them as brands and them working together will bet on. This is how it's been in the world in technology always. Even if you look at like you know, a steam engine, you consider steam engine initially when it came it could explode, it could be dangerous. And over time they put valves. It kept on improving for 200 years it kept on improving. Now when you think about steam engine, you think about the whole thing together, you don't think about it separately. That's how these things will evolve as well. So we are confident though hallucination is there because of how these companies that invest in it will operate. They will push these things along to make them more useful. That is one large trend. The second one I will share and stop is especially with AI agents coming, it is like in the future of organization research. When I was writing about it, I wrote about this like six, seven years ago. I called them commercial agents at the time. It's a bad name. So the thing was it was predictable that as technologies comes, they will try to create things that do automatically seek opportunities and work. What I see out of that is because of AI agents. While everybody's talking about new processes and things which will all be there, I think the larger opportunity will be using technologies to find inefficiencies and new opportunities that exist in the economy today and going and solving them much cheaper and much quicker than ever done before.

It's definitely possible, right? Yeah. When we talk about agentic AI, just how accessible the technology is and just how much cheaper it's getting, what we can accomplish by being AI first is, is boundless, I'd say. But you know, as we wrap Russia, we, we covered so many things. But what would you say is the one most important takeaway that you think our audience should have from this conversation? When it comes to being an AI.

First company, there are very, very few moments in our lives when we will have something as big that occurs in front of us. Right. Whether it is in, in culture, politics, technology, economy, so few, so few generation at so few times will ever have that we are at that stage. So as you know, we are on the Thursday before the holidays. Like you know, as people spend time over the holidays and starting the new new year, spend some time. Whatever line of work you do, try to play with gen AI tools. They are free and cheap with no preconceived notions. Be open to it. Don't stop with the first question and say it's not doing it right, just keep working with it. Just keep trying it. If you don't want to do it for work, pick a hobby, try to do interesting things like that. Do it as a family, do it with friends, do it co workers, keep doing it and you will find like how things you can do better in your personal life and in your work lives. I would highly suggest that and watch for it. There will be more and more developments. And this, what I'm suggesting is not just for only people at the front line or people at the, you know, doing the coding work. It is if you're a CEO or if you are a new founder or an investor, anybody. It's not only for content generation, it's for anything that we do. You can help it for every aspect. Pick a task. One thing I'll give you is if you have some project something coming is how can you be as an individual, be AI first for that one task and do it for some time, maybe a week, and then see how it plays out. Then you'll build more on it. And second, download Nourish. It's already available. Start 2025 on a good footing. It's fast and easy to track, much, much faster. Less than 10 seconds to track what you eat. And it's free. We'll be going to subscription model soon. Till the time whoever gets in now, it's a lifetime subscription is free. Please feel free to download. It's on the Apple App Store.

All right, well, thank you, Rajesh, for walking us through this. I think even hearing, you know, your own personal journey of how you've done this and, you know, with your background, I think is going to be very helpful for all of our business leaders that are listening and learning along with us. So thank you so much for joining the Everyday AI Show. We really appreciate your time.

Thank you for having me, Jordan. I enjoyed the conversation. Have a wonderful holiday season. All right.

And hey, y', all, that was a lot of great information that was just, just shared. If you missed anything, don't worry, we've got you. If this was helpful, please make sure to subscribe and leave us a rating on the podcast. Whether you're listening on Spotify or Apple Podcasts, as well as sharing this, if you're listening online, LinkedIn tag your friends, they need to know. But we're going to be recapping every single thing in our newsletter. So if you haven't already, please go to your Everyday AI. Going to keep you up with the AI news. If you missed it today, don't worry, we got it for you. So thank you for tuning in. Hope to see you back tomorrow and every day for more Everyday AI. Thanks, y'.

All.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going for a little more AI magic. Visit your everydayai.com and sign up to our Daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.