Investing in Generative AI with BGV Portfolio Founder Deepak Sekar | Ep 77

So we had to build our own AI voices. And remarkably, we have the best AI voices in the market, and we developed some special secret sauce to do that. And everyone who plays the VC game, whether it's a founder or whether it's a VC, they all need to think about making it a win-win. Often people get too greedy and just think only about themselves.

Hello everyone, and welcome back to Seed to Harvest, a next-generation venture podcast. I'm Paige Finn Doherty, the founder behind Genius Ventures, a venture capital firm investing early in the future of work and the future of play. I'm very excited to be joined today by Dr. Deepak Sakar, who is not only one of our portfolio company CEOs, but an investor in our second fund. Deepak is the CEO of ProfGym, a generative AI company turning text into interactive videos with avatars, which I'm excited to dig into in today's episode. His previous company, Chowbotics, sold to DoorDash. And Deepak, fun fact, holds over 225 patents, which as we discovered before the show, is more patents than Elon Musk, and is an avid mountaineer. Deepak, welcome to the show.

Thank you, Paige. Good to be here.

Yeah, I'm excited to chat today. So my first question is, you recently posted on LinkedIn about getting back into mountaineering, and I'm curious, what have you learned about leading a company from your mountaineering adventures?

Well, one thing about mountaineering is I typically go mountaineering on Saturdays every week. One thing about it is you're struggling every Saturday, fighting against odds and trying to get to the summit and come back safe in one piece. And it's a lot like a startup. You're fighting one battle after another in a startup. And so you kind of hone your— startup grit, I like to say. So that's one reason why I like mountaineering. It's an interesting challenge and it keeps people fit as well, especially once you get product-market fit and you start selling and got to sell a lot of people. Believe it or not, being fit helps. I've been— in my previous company, there were periods when I put on weight and there were periods when I lost weight. And I found when I had lost weight, I was able to sell a lot more. My numbers actually got better. So people make decisions based on the energy levels they perceive in the person who's selling. And they make decisions on, uh, they make decisions based on how others look. It's sad, but true. And the only way to handle it is to get fit.

That is definitely not a response I expected, but I'd love to hear more about what sparked your original idea for Prof Gym.

So like you mentioned, my background's inventing. I like coming up with ideas. And a lot of the ideas I come up with are in response to things I face every day in my life. So with my previous company, we built robots which made food. And that was in response to, you know, every time I cooked at home, I used to think, hey, it's so boring and mundane. There must be some way to do it better. So I started thinking about a robot which would cook my dinner.

Yeah.

And that's how that company got started. And so the company started out that way. And then we started making robots, which made salads. And when it got acquired by DoorDash, that company, we had more robots in the market than all other food robotics companies put together. And that company fed a lot of people good, healthy food in the middle of the night because robots don't need to sleep. A lot of hospitals are using it. Late in the night. With Prof. Jim, what happened was I saw my kids. I have 2 daughters. My younger daughter was 5 years old and she was sitting through these Zoom classes. And asking a 5-year-old to sit in front of Zoom 7 hours a day and listen to a teacher isn't really a great idea. My 5-year-old was getting really bored.

Mm-hmm.

And so I actually sat with her in a room to see why she was getting bored. And I saw her teacher teach and The teacher is this really charismatic person in person, but on Zoom, it kind of reduced her to a talking head. You couldn't feel her energy. You couldn't see her hands clearly. Not all teachers are great at lighting up their room properly. And so it wasn't a very good experience. So I started thinking, hey, there must be some way of using AR to make it easier for teachers to teach. So I started looking into it, and I called a couple of my friends and asked them if they wanted to tag team and work on this project together? And they said, yes. And so we started looking at AI, and this was before the term generative AI was invented. So we started looking at it in 2020. We filed a lot of basic patents on the technology on using these AI teachers to teach. And now everyone thinks AI is a hot area. And thankfully the US Patent Office has given us a lot of basic patents on it.

Yeah, I think that was one of the most interesting things that I learned from Fund 1, because we invested quite early in the space, but it didn't really have a name back then. I was like, oh, this is like really interesting technology applied in a novel way. And then it kind of post-AI becoming a hot space, I was like, oh, a generative AI company. There we go. Which I think has been a really interesting transition to Watch the space grow. I actually, I would love to like hear some of your thoughts on like how you've seen the generative AI space grow as you've been very early to it.

Yeah, so with a lot of ideas, you want to be early to market. With most spaces, once they become hot, you have a lot of people jumping on the bandwagon. Yeah, AI has been the same way as well. So some of the new people I'm seeing entering the market They don't have patents, they don't have much differentiation. They just call GPT. And some of the products I'm seeing with AI, I'm looking at it and I'm thinking, hey, that's something I can build in half a day. Are you sure you should be building a company on this? But there's hundreds of those companies where someone can just, you call GPT and build that product in half a day. There's no differentiation at all. And they're all getting funded, which is kind of— weird and crazy, but a couple of years from now, a lot of them are going to be copied out of existence and they're going to see low margins. So my advice to AI founders always is make sure you've got differentiation, make sure you've got some sort of moat around you. The moat could be patents, the moat could be network effects, could be some expertise you have, but you need to build some differentiation is what I always tell people.

Yeah, I think that's, I've actually, we've like diligenced some companies together and I remember you sent that over in an email. It was like, yeah, I could build this in-house in half a day with my team. And I was like, oh, okay. I feel like a lot of other companies are going to answer it that way. So it's been really helpful to have your expertise in that perspective. I would love to actually hear more about what it's been like working with me and behind Genius Ventures over the past year and a half since we met.

Sure. I kind of remember the first time we met, we met you and our products meant for younger people because we are doing education stuff, right? So we built products for kids in school. We built products for people in universities. So our customer base is anywhere from 10 years old to 25 or so. There's some people who are older who use our products, but 70% of our customers are of age 10 to 25. And so it helped for us to have an investor who is of that age group, and you kind of got it immediately. And it's rare to find people who get your idea immediately. A lot of people, they don't have enough expertise in the space, so they're always doubting themselves on whether it's going to happen or not. But you're like, this is going to be great for everyone in my generation. You know, I get it. And when I saw that, I was like, Paige has intuition. And when I think about early-stage investing, a lot of it I think is based on intuition. You got to know what the market wants, got to know what the customers want, and you need to be able to judge people well. Those are the best early-stage investors I've seen. And I thought, You know, you, the fact that you got it immediately and you had good insight on the space, I, I thought, hey, this is an investor I wanna work with. And so that's how it started and it's been a great experience so far.

Yeah. Yeah. It's been, I think one of the most rewarding parts about being an early stage investor is like watching the companies that you have early conviction in grow, like seeing Prof Gym grow. One of my, actually one of my favorite things that you do is send your investor monthly updates with using the Prof Gym technology. So. We'll get an update from like Sacagawea, we'll get an update from Abraham Lincoln, or we'll get an update from your avatar. And just in the past year and a half, like the progression that you've had in developing very like human avatars has been like so insane to go back and like watch some of the earlier updates. Um, and I, I think like the, the product obsession is Has just been like so incredible to see executed across that moment of time. So that's been really cool.

Thank you. And Abe is great even in sales meetings. So I still remember we met the CEO of a pretty big textbook publisher just a few weeks back and we had Abe show up and say, hey, I really like the way you're building this company. And he found it quite interesting to have Abe Lincoln compliment him. It broke the ice. He gave a smile. And I like saying when someone smiles, they're sold.

Sure.

So yeah, it breaks the ice and it's interesting using all these famous characters. And we also, it's also fun to work with all these famous historical characters. Like I personally had read Jane Austen's books a while and we actually recreated Jane Austen. And so we did some research on how she would look like and we tried to mimic her appearance. We recreated Gandhi recently as well, so I had to read books about Gandhi to understand what kind of voice he had. We actually took his voice and cloned his voice, and we tried to figure out how he would look like and what he'd say. So it, it, I don't think of it as work. It's kind of fun creating all these characters and having them teach stuff.

Yeah, I feel like the sense of play is so prevalent in how you think through building Profit Gym, which is really cool. I would love to hear more about maybe some of the challenging technical obstacles that you've overcome as you've been building Prop Gym, like things that were challenging. Yeah. That just like stumped you and you got through.

Yeah. So with Prop Gym, one of the things is it's a very broad company in terms of skillset because we've had to develop new AI voice technology because all the existing AI voices in the market they can't keep students' attention for a while. So we had to build our own AI voices. And remarkably, we have the best AI voices in the market, and we developed some special secret sauce to do that. And the entire company is just focusing on voice AI, and the fact that it's a portion of our company, we have 3 engineers work on it, and we have the best voices in the market according to our customers. So that's pretty exciting. The other thing is, besides just AI voices, we had to learn all about computer graphics and figure out how to get these avatars more realistic and all. And so we had to learn a lot of new skills there. So I had to learn a lot about this myself. Thankfully, we have a professor from Stanford who's actually won 2 Oscar Awards. He's one of our advisors, and he's been helping us ramp up on that. And now we've gotten pretty dangerous. We can do some really cool work with these avatars now. And we added a few people on the team as we raised more money. who are some of the best game designers in the world, and those people have been doing amazing work. So we had to learn skills there, and then we had to learn front-end work with designing websites and coming up with these web portals. So what Canva has done over 6 to 8 years, we had a team of 3 people kind of do in 6 months, which is pretty remarkable. And besides front-end work, we also had to— we have a team focused on large language models. So the breadth of technology we have has been like huge and learning all these new skills, it's been so much fun, but it's also been pretty challenging to learn so many different things about so many areas.

Oh, it's, that's so cool. I feel like some of your work with like face expressions was super interesting to see as well, like early days. And like hand motions, like how you think about modeling those. I, it was, yeah, it was, it was like super, super cool. I'm just like really excited about what you're building. I'm curious to hear like, what are some of the lessons that you've applied building Prof Gym after having sold a company?

So the first time you do a company, you're kind of learning on the job, uh, just to be honest, and everyone knows it. So the second time. is when you actually leverage your previous experience. So I made a lot of mistakes in my previous company now that when I think back, one of them was I tried to do it on my own as a single founder, and it was a turnover cost. So this time around, one of my big learnings was get multiple people to share the journey with. It's just too much work to do it on your own. So I got a couple of really good co-founders, and the 3 of us We've been working together so well, so it's made my life a lot easier. So that's one big piece of learning. Another big piece of learning, I think, is to have good diversity in the team. Last time around, we had an all-male board, and a lot of our execs were men in my previous company. And this time around, it's a lot more balanced. 50% of our employees are women. 50% of our managers are women, and we have one co-founder who's a woman as well. And our board is like 40% women as well. And we find having different points of view really helps. Like, for instance, if I'm gonna create an avatar of Jane Austen, if you have 3 guys build it, it's gonna suck. And so, so the fact that you have, you know, people from different backgrounds all working together, There's only so much about makeup a guy can learn about to apply to an avatar. So we have— normally when you have to design a woman's avatar, we have a woman designer, a 3D designer who works on it, and it looks so much better compared to a guy doing it.

Yeah.

And that's just one example. But even with decisions we make at the board level, right? If you have 5 guys in a room, A lot of them often think the same way. But if you have a couple of women, you don't have 5 people all ganging up together going one direction. You have people with a different perspective adding some thoughts as well.

Yeah.

So I found having that diversity has really helped this time around. And the interesting thing about diversity is it starts right from the beginning. If you have one co-founder who's a woman, She can attract a lot of good women into the company. And if you have half the managers are women, a lot more women feel like joining. And if you have half the board or 40% of the board is women, everyone's got role models to look up to. So I think that was one piece of learning I thought was quite valuable. And if I had to pick a third piece of learning, it was, it's the mountaineering you brought up. The last company I was working 7 days a week, 20 hours a day. And I did it well for 3 years, but in year 4, I started burning out. And so this time around, I've tried to take 1 day off in a week. So I try to take Saturdays off to go hiking or mountaineering, and that brings me some calm. And so it's good to get that break as well. And so there's a lot you can accomplish with Working 6 days a week. You don't have to be on 7 days a week, 20 hours a day, 6 days a week. You know, 18 hours a day is not a bad thing.

I'm so inspired by you, Deepak. I'm so happy that we get to work together. Um, I, I thought I'd take us back to our first call. So you gave, um, us some perspective on like what that was like for you. And I wanted to talk a little bit about the things that I noted on our first call. So my first 2 words that I put down were customer and then in all caps, obsessed. And I think that that has truly defined like one of the reasons we were so excited to invest in Prof Gym. And also I think one really interesting thing that was kind of like insider baseball was that you had like a deep understanding of VC math. You talked about like recycling and QSBS in terms of an early exit, and we're very like thoughtful about the business as a product and a focus on building a diverse team, which was really interesting because I feel like there's a couple different layers that you can look at a business. And I always say I like backing founders who can kind of look at the business with a microscope, but also with a telescope. And I feel like you had this really interesting perspective of the business as a product. that was very prevalent through you talking through like VC math that I had never heard any founder talk about on an intro call before, which was really cool. So I, yeah. Do you want to, do you have any advice for like founders going through that journey of like fundraising at the early stage?

Yeah. So I think one important thing is to make it a win-win for everyone.

Yeah.

And everyone who plays the VC game, whether it's a founder or whether it's a VC, they all need to think about making it a win-win. Often people get too greedy and just think only about themselves. And so even things like valuation, right? You could be greedy and try to grab as much as you can, but it comes back to hurt you in the next round. And if the VC doesn't get a good ROI in terms of returns, The VC is not going to invest in you next time around. So with my previous company, the investors who, you know, the investors who did well, they put money into this current company, you know, and I think of it as a relationship. If I get them a good return this time, I don't have to worry about fundraising for my next company anymore. So I think everyone's got to think about it that way. There's some VCs who try to squeeze every penny to show a good return in the current fund. But they're never going to have the founder work with them again in a future company. And normally the founder gets better in the next company and the founder's going to get a better return. But if you try to squeeze every penny and, you know, don't give a founder a good ROI, the founder's not going to work with that VC again. So we've been very fortunate this time where every VC we work with, they understand it's got to be a win-win. You folks have been great. We've been fortunate to work with other investors who are that way as well. And it just makes for a more harmonious relationship as well. It's more fun to work with people like that instead of people who try to make it a win for them and try to make it a lose for others.

Yeah. Yeah. Well, one thing that I like to do with guests is ask if you have, or yeah, what questions that you might have for me.

Yeah, I guess one question is I'd love to hear where you see the future going. So as VCs, you folks have to need— have to think 5 to 10 years out, right?

Yeah.

So when you're thinking about portfolio companies for the next 5 to 10 years, what's your crystal ball saying?

Yeah, well, I was, I was just gonna say, I don't have a crystal ball, but I think that Some of the things I'm really excited about, especially in today's market, I think there's some really interesting opportunities happening in consumer technology right now, or B2B2C. I think one of the things I think a lot about is like we're past the day of paid ads working really well. And so I think there has to be like some unique distribution advantage to companies that will grow really quickly over the next 5 to 10 years, whether that's like a new social network that launches and they capitalize off of that. Or they're building like deep brand equity with a very specific set. I think Megan Loist did a really interesting newsletter piece on the 80/20 model, which is basically that like 80% of brand equity is like built through 20% of the creators that you work with. So if you look across like any company, take for example, like iJustine and Apple, she's become like one of the foremost Apple influencers, which I think is really interesting to look at. So trying to identify in parallel, like which creators will define culture over the next 5 to 10 years and have significant staying power. And then what brands or companies do they choose to partner with or champion, I think will be a really interesting tell as to companies that stay around for a long time. I think we've seen this in the B2B world with a lot of Twitter creators as well. Um, if you look across like newsletter writers, like Um, Packy McCormick, uh, or Turner Novak. I think they've had, they've like gone into specific niches, whether it's deep tech or grocery delivery, based on the feedback that they've gotten from building an audience online. So I think a lot about how building an audience can give you additional feedback into unique consumer segments and Then once you built that, it takes a lot for you to be able to champion a brand. Like you have to be like truly excited about it to be like, all right, I'm gonna go out on a limb and risk like my authenticity as a creator to be able to champion something. So that's something I think a lot about in terms of the future of companies I'm excited about is evaluating like who I think are up and coming creators in across like B2B and consumer. And then some of the trends that I'm actively following, I, I think there's like an enormous shift happening in 2 places across the arc of the family. So the first is like Gen Z as parents, like what is that gonna look like? What technology are we going to use? Like, I think that's gonna be really interesting to see how that plays out. And then the second is a rapidly aging demographic, especially in the US, with a senior population that is like ready to adapt technology to expand their independence and agency into their like silver years. And then also how technology plays a role in the caretaking economy as more people just like in their middle age start to care for their parents. So those are 2 of the things that I'm like really excited about is the arc of the family and then looking at the different like up-and-coming creators as a signal. Yeah.

That's interesting.

Yeah.

Yeah. I'm thinking of how AI and robotics could impact things like that because you would imagine with AI becoming so good, some of the speech-to-text implementations now are able to understand accents very well. You can talk. These text-to-speech things sound like real people as well. So you can think of a future where you have robots which can talk and listen and can walk because self-driving cars are here in San Francisco. So robots can navigate as well.

Yeah.

So maybe the older people you're talking about are gonna be surrounded by robots helping them around the house. A good chance that could happen. And some of the science fiction movies we saw 10 years back where You, you had companies building robot fleets, helping people as servants and helping older people. You know, we are not too far away from that. I don't know if Tesla will be a winner there or Apple would be a winner there or who'd be a winner there. I know Google is working on it as well, but the next 10 years you might see some interesting technology there, I would imagine.

Yeah. Yeah. I think that, that like intersection is really interesting. Actually, one of our investors and advisors at Behind Genius Ventures helped build the wearables business at Qualcomm. And he, one of the things that he brought up was that the, their like elder tech was one of the fastest growing segments because of that. As you combine a population that's had familiarity with technology and a willingness to adopt it with like a very large generational wealth transfer and a, an urge for longevity and agency and independence, I think it's going to be really Really interesting to see how that plays out. So I'm following that pretty closely.

And you can think of devices now that you've got watches which measure heart rate well, they measure blood, the blood oxygen levels as well, and you can measure pulse rates too. And so you could think of a future where you get a lot of data about each person simply from the watches they have, and you could even get that temperature and things like that. And There might be a lot more health solutions which are offered online because you have enough diagnosis with just watches and other such things. And so you could just stream the health information to the doctor who's online, and you'd have a lot more things being done through Zoom calls. And you could even think of a future where you have AI looking at all this data and giving predictions on what someone's ill with. And the doctor looks at it, the human doctor looks at it.

Yeah.

And uses that as a first-pass diagnosis and then sees if it makes sense or not and probably gives a better diagnosis than them just, you know, guesstimating what the illness is based on their gut feel. So there's a lot of interesting things that are going to happen in the next 10 to 20 years. So it should be a fun ride.

Yeah. Yeah. I think it definitely is. So I feel very grateful to be in the position I am today to get to back early-stage founders building at the future of work and play side. Any last thoughts before we hop off today?

So in terms of last thoughts, uh, I guess you've got a bunch of entrepreneurs following your podcast. Uh, the one piece of advice I would give is, uh, to try to build a things which are 10x better than what's there today. A lot of people who start companies nowadays, they just do something incremental, which is a small change over what's over there. And that sort of company building just results in people copying you after a few years and you not having enough of a moat. So just aiming for 10x, something which makes society better, which makes the world a better place, you know, considering that For starting a company is pretty important. So I guess that's, that would be my advice.

I love that. Well, Deepak, thank you so much for joining us today. And if you want to get in touch with Deepak, I will drop Professor Jim's information down below. And thank you so much for coming on.

Thank you for having me here. It was good chatting with you, Paige.

Of course. And a very special thank you to Seed to Harvest podcast editor Tate Doherty for his amazing work on this episode.