Vijay Reddy, partner at Mayfield, talks about the latest innovations and trends shaping the artificial intelligence (AI) landscape. He shares his venture capital journey and describes Mayfield's unique focus on cognitive plumbing and Cognition-as-a-Service (CaaS), unraveling what these concepts entail. Vijay also offers practical advice for startup founders keen on tapping into the myriad opportunities within the realm of AI.
Vijay Reddy, partner at Mayfield, talks about the latest innovations and trends shaping the artificial intelligence (AI) landscape. He shares his venture capital journey and describes Mayfield's unique focus on cognitive plumbing and Cognition-as-a-Service (CaaS), unraveling what these concepts entail. Vijay also offers practical advice for startup founders keen on tapping into the myriad opportunities within the realm of AI.
In this episode, you’ll learn:
[1:24] From a 10th grade circuits and chips enthusiast to technology executive and founder
[5:50] Vijay’s evolution as a venture capital investor fueled by technology curiosity
[13:41] Debate about AI being a good or bad thing - it’s not worthy!
[15:04] Get these 3 things right for success with an AI-first startup.
The non-profit organization that Vijay is passionate about: Upward
About Vijay Reddy
Vijay Reddy is a partner at Mayfield with over a decade of AI and deeptech investing experience. Prior, Vijay was an investor at Clear Ventures and Intel Capital. He witnessed AI's rise from its inception, and with a founder-first approach, he has invested across the AI stack, backing companies like SambaNova, DataRobot, and BabbleLabs. He began his career as an entrepreneur, co-founding a startup after leaving his PhD program, and has held senior roles at Broadcom and Intel.
About Mayfield
Mayfield is a Silicon Valley-based venture capital firm that invests in AI-first companies
at the Seed, Series A and Series B stages. The firm has raised 20 U.S. funds since 1969, and currently has $3 billion under management. Mayfield takes pride in an investment team that has a founder DNA and operates from a shared set of beliefs and partners for the long term with entrepreneurs pursuing big ideas. Companies in Mayfield’s portfolio include: Outreach, seekout, alchemy, auradine, chemix, crunchbase, Gutsy, LexCheck, OwnID, Qwiet, Vijil, VERSA Networks among others.
Subscribe to our podcast and stay tuned for our next episode.
Founders are a different breed of people. There's very high chance of failure at the early stage and you have to be irrational to be a founder, unless you believe in yourself and the technology. So it's the kind of people we are attracted to.
Gopi Rangan: You are listening to The Sure Shot Entrepreneur - a
podcast for founders with ambitious ideas. Venture capital investors and other early believers tell you relatable, insightful, and authentic stories to help you realize your vision. Welcome to The Sure Shot Entrepreneur. My guest today is Vijay Reddy. Vijay is a partner at Mayfield and he focuses on artificial intelligence.
We're going to talk to him about cool things that are happening in AI today. We're also going to talk to him about his journey as a venture capital investor. What are things that [00:01:00] excite him? What kind of founders he likes to work with? And why does he say no sometimes or probably often? And what are some reasons for that?
Vijay, welcome to The Sure Shot Entrepreneur.
Vijay Reddy: Gopi, thanks for having me. It's a pleasure.
Gopi Rangan: Tell us about yourself, starting with where you grew up. You were from Bangalore, and then eventually you migrated to the U. S. You were a grad student, and you were an engineer, and eventually became a venture capital investor.
Let's start with your journey in Bangalore.
Vijay Reddy: Yes, I grew up in, uh, Bangalore. It used to be, uh, known as a garden city, small city. And actually, I grew up, I was born in Bangalore, and I, that was close to Bangalore. My grandparents were agriculturalists and my dad was an engineer. So I, I was always interested in business because my dad became a businessman.
My life has always been around real estate construction and engineering. And somewhere around the 10th grade, I was doing a project and I really liked the circuits and chips. So I didn't know what I want to do, but I kind of liked the [00:02:00] circuits piece of it.
And I did my engineering and electronics and I did a stint at IISC where we were doing some, again, just kind of chips related. So I got really fascinated. Came here for my integrated master's PhD and, uh, I just liked the RF circuit design and that's a theme that carried always in the back of my head is that maybe I should go back and do real estate.
in India at some point, uh, because I was, I had the rosy picture of that because Bangalore was booming, right? It's the fastest growing city in Asia. It felt like it's very, very active. And so, and it was very entrepreneurial also. So those were the things which keep pulling me back. And so in my second or third year of PhD, uh, my PhD was mostly integrated RF IC chips, which is basically chip design for your radio, for phones and Bluetooth and your wireless and so on.
And I wrote a concept paper saying, "Hey, [00:03:00] look, we need to do two different phones, one for Japan, one for U. S. because you had this CDMA and GSM and you needed different circuits for both because one couldn't handle multiple. And if people remember, you have to carry two different phones or two different things.
And so there's a bunch of those postdocs and PhDs. We were working on a technology which could go across different bands and we could build this thing called power amplifiers or which could be multi band multi mode and it was a very interesting project. We wrote a concert paper, won the best concert paper award and wrote a business plan.
We won the first place and 15, 000 and we pitched to VCs and it was very exciting to the point where we had some grants from I think it was Connexent and Broadcom for the sponsors. So a couple of us, or at least I quit given it was so exciting and a couple of the postdocs are still continuing with this and they went back to Taiwan.
My professor was from there. So it was, it was really [00:04:00] exciting to the point that I, that was my first dabbling into venture on my own or at least into startup ecosystem. Then I joined Broadcom in the early Bluetooth Wi Fi group. We grew very, very rapidly. We got the Apple design win and we got a bunch of others.
We were building this combo chip, Bluetooth, Wi Fi. And it was kind of like a chat GPT moment back in the day because chips were built on Silicon Germanium RF and people are porting that onto CMOS. And, now you could build Wi Fi, you could put all kinds of radios on CMOS and it was very, very smart team.
I did that for maybe a couple of years and then did product and biz dev. And somewhere along the way I thought, "maybe I should think about entrepreneurship again. This is a good ride at Broadcom, but how do I go about thinking about business?" So I went to business school and I had two focuses. One was finance and one was real estate.
And I thought I'll just try this a little bit more. And during the treks I went to India. The ecosystem had changed [00:05:00] dramatically, the private equity ecosystem, especially around real estate and finance. And it was not what I wanted to do. So I wanted to come back into tech and again, being a broadcom, it was either entrepreneurship or joining a fast growing tech startup.
And I got into this program called the Accelerated Leadership Program, which is a program at Intel where they pick one from each top 10 business schools. And they groomed you to be a GM or VP. After six months getting into Intel, I was like, okay, that's not what I want to do. I was like, that was not my calling card to be a VP, GM of a business unit.
Uh, I liked tech. I liked entrepreneurship again. So I was like, okay, after six months I realized it was not my cup of tea. And luckily by then I got into Intel Capital and maybe that's the second accidental part of the journey into venture.
Gopi Rangan: Fascinating. You were a nerd very early in your life. You geeked out on circuits and engineering as early as like the 10th grade.
And you had [00:06:00] always this itch for real estate, but you came back to tech every time. And you actually dropped out of your PhD program to start a company. You built your career as an engineer early days and then eventually switched to venture capital. Why do you like venture capital? What is most exciting for you?
Vijay Reddy: Yeah, so I think the journey to venture capital was also kind of tech focused, right? So in 2013, when I joined Intel capital, I was given a project of go find computer vision and build an investment strategy for Intel around computer vision. And back in the day, computer vision was about. It was about how do you get different kind of imaging cameras.
Intel wasn't a mobile at that point, and it was all about race to get to five megapixel, six megapixel computational photography. And just around the time AlexNet had come out and it's basically neural networks for the first time surpassed human accuracy. So if you take a ImageNet challenge, we could look through a million images and [00:07:00] try to find which has, how accurate can a computer vision system be compared to humans.
And we just crossed that barrier. And I remember when I was doing my PhD, my roommate was a computer vision PhD student, and he had built a ton of models to look at human faces and across different camera lenses, uh, ultraviolet infrared. And his entire PhD was disrupted with this CNN because now I don't need to build features anymore.
I could just train on vast amounts of data and I could just use compute and I could build this model, right? So this curiosity of, Hey, look, what if I could stretch this further? What if cameras can see better than humans? What, what would happen, right? That was the investment thesis. And luckily for me, I think Intel was paying attention to that.
And we said, "Hey, look, what if cars could see things better than humans can. What is the big possibility there?" It was a seven to eight trillion dollar market that could open up. [00:08:00] What if satellites could see and we could actually measure oil reserves. So you had satellite imagery, medical imagery, and you could go on and on and on with that breakthrough.
You could find lots of pockets which were exciting, right? And to your question about venture capital, for someone I wouldn't classify as a nerd. I was more like a curious and for someone who is interested in the business side of things and the people side of things, it's a natural fit because things are changing so fast and founders are different breed of people. There's a very high chance of failure at the early stage and you have to be rational to be a founder unless you believe in yourself and you believe in the technology. So it's the kind of people we're attracted to, someone who's contrarian, someone who can take the risk.
And back in the day when deep learning was new, the kind of founders attracted were people who are visionaries. And, uh, those are the kind of people, even today, we like to engage with.
Gopi Rangan: I see your, [00:09:00] uh, desire for learning and your heightened level of curiosity. And I think venture capital is probably the best field for someone like you and for me as well, like, similar reasons.
We get into new topics that the world hasn't seen before. And we work with founders who are largely irrational. They spend so much effort into becoming the world's best expert in that topic. And we get to hear their vision for the future. Now, some of these work, some of these don't, and when it becomes really successful, it really changes the way we live.
I see that the way you developed your thesis for computer vision really rooted your foundation in venture capital. And then you went on to do many other investments. You are at Intel Capital, you are at Clear Ventures, and eventually now you are at Mayfield. Can you give a quick description of the fund you manage at Mayfield? What kind of investments do you focus on?
Vijay Reddy: Yeah, so I think to answer this question, maybe I'll share my evolution as an investor. So it started off with technology curiosity. What is the next cool thing or [00:10:00] next big thing that people can build. So initially it was looking at people who can build cool stuff.
And as you make more investments, you learn that you need people who can't just build products, but actually build companies. And eventually you learn that it's not about the technology or the product or the company. It's about the people. So it's been a evolution of my investment journey is to focus on the people and not just on the product or the technology.
And that's been a learning experience for me. And when I was looking at where to join I think one of the nice things about Bayfield is where people come first, right? We focus on the jockey as long as we are focusing on the big team we heavily rely on picking the right founding team who can go and execute on this journey.
So the people first philosophy is what resonated with me quite a bit. We tend to back people, but we do make sure they have the same values as us before we invest. So I think that's the first bar for us is to be [00:11:00] people first. And we don't do a lot of spray and pray. We're very focused because we think in the, it's a craft and we have a lot of resources.
And so once we find the right person on the team, we spend a lot of time working with them. So we don't have SSH or unless it's repeat entrepreneurs here. All of them have been operators before. Many of them have run $100 million ARR businesses. So we tend to focus on how do we add value from day zero and a lot of companies at inception stage.
And the results have been over 500 investments, more than 120 IPOs and more than 250 MNAs. And that's a pretty high hit rate, right, for the kind of investments. That's because we do a few deals and we bet on people and usually that's a winning formula unless the markets change or something dramatic happens.
In the AI space, it's opening up so many frontiers similar to what happened 2014. I think if you look at AI, even just the last three years, right? We [00:12:00] have logic, reasoning beating human accuracy. We have them beat, Gemini just beat in November benchmark on MM, the multimodal, the models beating some of the assessments out there.
Uh, for example, AI can do better than a legal person or can pass AP bio exams. So there'll be a lot of breakthroughs in AI, which open up a lot of different fields. So the two questions we asked was, what is Mayfield? We're a 54 year old fund. People first. We announced a billion dollar fund.
So 250 million is an early stage AI start, which I'm part of. We have a flagship fund and a spring fund. Most of these are inception stage companies. Seed A and B is our sweet spot. And we tend to invest across planetary health, human health, enterprise SaaS, and infra. So in the AI start itself, the themes are very AI focused, but we have this theme around cognitive plumbing and cognitive assistance service.
I'm happy to chat about that. We're extremely excited about both those spaces and [00:13:00] I'm sure a lot of founders would find it interesting as well.
Gopi Rangan: Well, thank you for sharing your journey, your evolution as an investor, how you were initially focused on the technology side of things, the product and the market.
And now you've kind of evolved to focus on the entrepreneur and the jockey. You bet on the jockey, you are people first, and the market and product and other things come later. Let's talk about your areas of focus now. Cognitive plumbing. What does that mean?
Vijay Reddy: So we have two main themes around cognitive plumbing and cognition as a service.
We call it CAS. So as AI has evolved from let's say object recognition to like things which are more cognition kind of capabilities and as we look at things beyond cognition of how do you go from idea to actions. you need a different kind of plumbing to make cognition available to the masses. And so on the lower layer, we have the hardware layer, right?
So it's chips, whether it's ASICs, CPU, GPU, networking. [00:14:00] On top of it is what we call the infra software layer. Think of that as the networking storage security layer, which is very interesting for us. Then you have the integration gateway and the models on top of it. These are the foundation models. So that forms the base of the cognitive cloud.
And on top of it is the data, middleware and apps. And then we have this notion of how can agents help humans as co pilots and going forward be co workers to these humans. And so we call that the cognition as a service or CAS. How we went from IaaS to PaaS to now we're going to the CAS layer.
Gopi Rangan: Very interesting. You look at the two sides of it starting with the infrastructure and then all the way up to Before we go into more of your areas of interest in AI, let me ask you this more controversial, popular question. Is AI a threat, or is it a good thing? Net net, long term.
Vijay Reddy: I'm more [00:15:00] in the optimistic camp. I know that there's a debate, I think it's not worth the debate.
Look, historically, We have found that technologies, people tend to use them more effectively and the more you use them, the more you can expand the market and it can go all the way back to cotton farmers or factories or even more recently computers. I think it has a massive benefit to the society and people will evolve to use it as a tool just like a calculator or a computer.
And any short term concerns will be allayed by productivity gains, right? We're going to see companies where fewer people are going to build more, but end of the day, productivity will increase. And I think people interpret that as people losing jobs. So for example, you can get $100 million ARR maybe a hundred people, maybe 10 people, maybe one person, right?
I would look at that as how people can build more with less than people who lose jobs. But there needs to be retraining, there [00:16:00] needs to be corporate level efforts to make sure a lot of companies will enable the workforce to learn and train. But that's the right discussion than losing jobs.
Gopi Rangan: You are clearly on the optimistic side of the AI debate.
Can you give a couple of examples of startups you've invested in? How did the conversation start? What do you look for in the first and the second meeting? I know we all look for basic things, but specifically related to the topic of AI and the way you focus. What are some things that excite you?
Vijay Reddy: Yeah, so I think a lot of times we're actually judging the people on the team. How do they present themselves? What is their vision? What are their values? Even though they're talking about technology.
So I think be honest, be truthful, be vulnerable if you don't know something. It's a good learning opportunity, right? So I think once you cross that people filter, it is about what is the big idea you have? How can this be a large company? So for any VC, you need to have big [00:17:00] outcomes and almost every company has to have a chance of a big outcome.
So start with what is the big vision, but then quickly come down to like, hey, how do you go start with a narrow wedge to go into that vision? And how do you build to that end goal, because it's a multi tenure effort. So going high, talking about the vision and then quickly coming down to here's what we're good at. Here's our unique view of the market and here's why we're the best team that can do it. That's kind of the three things we're looking for in teams. And then you can go into like, why does market care? Will they pay for it? And competition and maybe the product, right? So, uh, the sequence is first get the VC involved and explain to them that this is a big idea and then why you're uniquely qualified to go solve this.
Gopi Rangan: How is it different to build a company that's focused on AI compared to other sectors? What's more challenging for founders in this area?
Vijay Reddy: So I think AI has unique opportunities [00:18:00] and challenges. A lot of times people look at AI as a hammer at this stage, especially in the gen AI. They're saying, "Hey, look, I have a really good foundation model I'm building, or I have this really cool tech I built at a large company. I'm going to find a lot of use cases and going to pick one which actually resonates." That seems like a good logical approach of, "hey, look, I'm going to do market discovery and find it." I think a better approach would be, "here's what we found as a problem in the market. And here's a list of tools we have at our disposal to go solve it."
So think about it more from a customer standpoint and try to explain that to us of, here's what I found in the market versus, "Hey, I have a hammer looking for nails." So in the AI space, there's a lot of different unique nuances. A lot of companies which are building foundational technologies need a lot of capital, which 50 percent of that goes into compute. So that's the cost of starting some AI companies is much harder depending on where you are in the stack. So that's one class of companies.[00:19:00]
The other class of companies is, "look, we have a really strong AI team, but I want to go into legal tech." And then when you peel the onion, they don't have any legal experience. So that's one pitfall we're seeing is people are trying to go into domains where domain expertise matter more than technology.
In other domains, it's the opposite, where people have really good domain expertise, but they have no AI expertise. And that could have been possible in that vertical SaaS. But in many cases, if you're going into domains, you're competing with teams which have deep AI expertise, right? And I've seen that the company that's going to disrupt is someone with AI, but also without AI.
So we like teams where you have deep domain plus AI expertise, and not just hiring one junior engineer, but someone in the team. AI first teams are harder to find if they're going after vertical domains. But if they're going after developers or then then it does, you don't need to have the domain expertise, but you need to know how I take the compassion for developers or compassion for enterprise sales.
Gopi Rangan: So knowing which market are going after, [00:20:00] if you're going after enterprise customers, you better have some enterprise DNA and sales. If you're going after developers, that's a different DNA. So knowing who you're going after, what do you need in the team and building a team around it is very, very interesting.
So your advice to founders is that instead of 'if you build they will come' type of approach.
Don't just build something and hope for customers to show up. That might have worked in other sectors previously in the past years. But in this sector and I focus on the customer, focus on the problem, and know what problem you're solving and how important that problem is and how much they would pay to get that problem solved.
Making that focus on the customer first is more effective in building the business. And while you're building the business, make sure that the team has enough strength both on the AI expertise side and the vertical domain that you're going after.
Vijay Reddy: Yeah, I would say like being customer obsessed is not just an AI issue.
It's always been there, right? [00:21:00] I think it's just heightened now a lot of research teams who have AI expertise without domain expertise, right? I think the advice to anyone who is starting AI native is go back to basics, like how anyone would build companies even in the past. Focus on the customer problem, and then see how AI can uniquely solve it. It's always been the same.
Gopi Rangan: Can you give an example of a startup that has done this well?
Vijay Reddy: So, a lot of the companies, and again, since we focus on that, we tend to look for, if there's been a company which is looking at a persona, that's a developer. Right. Be developer obsessed and know exactly what they're solving for, right? We have a lot of companies are building co pilots for IT co pilots for SRE co pilots for red team blue team of companies and They all come from a customer problem than from a tech problem.
They could as well be solving any other issue, but they come from the SRE world. They come from the IT ops world. And for them, they've lived it and now they have new tools, which they can [00:22:00] go after the same problem. It's a very good founder market fit and that's very important for us.
And if you're in first time entrepreneur, you still have to have some compassion for the problem.
Gopi Rangan: I feel like you have a specific company in mind. What's the name of the company?
Vijay Reddy: So we just invested in one, which in a part of the AI start, right. It was called Vigil. It's in stealth. So a lot of those companies are in stealth at the seed states. So I'll use this one because it's, it's a recent one. This is a SageMaker team. They've built SageMaker over the years, the GM and the head of engineering, and they know what customers want.
They've lived this at Amazon, and now they're trying to solve this for startups to help protect models, to help evaluate the models, right? These are people who have spent their 10, 000 hours plus in other places, and now they're going and solving this for the rest of the world. So that's very helpful for us if people have empathy for a problem.
Gopi Rangan: I like the way you say it. It's the lived experience. It's not about number of [00:23:00] years of experience in an industry. It's about the first hand experience of encountering a problem and getting deeply attached to the challenge of it and absorbing all the things that are around the situation so that you can develop some conviction on how to solve the problem.
That is very important.
Vijay Reddy: And it's not that hard to find also, right? If somebody's at Juniper or Google or Facebook, they're living through issues every single day. And now you have Gen AI and suddenly wake up and say, "Oh, I can go solve this at scale for a lot of people who are having the same issues." That's the right kind of founder we want to talk to.
Gopi Rangan: How many companies do you typically invest in, in an average year? What stage is your sweet spot? You mentioned seed, series A and series B. That feels quite broad, but is there a sweet spot for you?
Vijay Reddy: I would say like a majority of deals across all funds has been at inception.
These are teams where we tend to be the first institutional check. At the AI start, it's more so [00:24:00] because we're in the seed fund and we tend to be the first check. Or as a syndicate, we invest in the first check. For example, in Vigil, we partner with Gradient on this. Otherwise, a lot of them are at series A, especially if it's vertical SaaS. If it's infra and teams we have known we tend to do it at the seed stage.
Gopi Rangan: And how many startups do you invest
in?
Vijay Reddy: It depends on the fund. So in the AI start, we're trying to do three to four, given the size of the fund. In the series A, I would say it's like 10 to 12, right?
Gopi Rangan: Okay, good. That keeps you quite busy.
Three to four AI startups at the early stage and the rest of the firm focuses on series A.
Vijay Reddy: Yeah, we have, we have a team of investors and that's all we do, right? We invest in companies. But relatively speaking, it's, it's, One or two deals per person, right? And it's not too many.
Gopi Rangan: What's your most common reason to say no?
Vijay Reddy: So one is obviously the people screening, right? We're very selective of who we pack. Some say we're very, very selective. So once we get a person we want to bet on. [00:25:00] Um, do they listen? Do they understand the markets? Can they build a team? Can they inspire people? Can they tell a story?
And you can tell fairly quickly if it's a pass, but if it's a yes, It takes a while to build a conviction.
Gopi Rangan: How long does it take from the first meeting for you to say yes?
Vijay Reddy: Typically we process deals really, really quickly. I would say like within a couple of weeks, but we would want to meet the team. We tend to invest heavily in Bay Area and the West Coast because we get to know the teams. We do well when we can meet the team often and we can help them more if they're local. So fairly quickly is the answer all the way from a couple of weeks, I would say, is where we can get to a quick decision.
Gopi Rangan: That's quite fast.
Vijay Reddy: Yeah, it's very, very fast, relatively speaking.
Gopi Rangan: Okay, we're coming towards the end of our conversation, and I want to ask you about your community involvement. Is there a non profit organization you are [00:26:00] passionate about? Which one? And I understand that mental health and childhood development disorders are some of your favorite topics.
Vijay Reddy: So I think there's two of them, right? So I'm part of the, I'm the board of director of Upword, which is one of the largest women's organization. There's not enough diversity, both in startup, because it's especially in the AI space and especially in the venture capital community. So that's one. The other is childhood mental health and DSM 5.
So I'm helping some founders build some companies around that space. That's a space which is very close to heart as well. And those are the two main things.
Gopi Rangan: On the board of a school called Achieve Kids and that focuses on special needs children who need a lot of support. The world definitely needs to care for children at an earlier stage when there are mental disorders and other challenges.
We as a society would benefit when we care for them at an earlier stage and help them develop some systems so they can thrive and be more [00:27:00] independent.
Vijay, thank you very much for spending time with me. Thank you for sharing your journey as a venture capital investor, how your thought process has evolved with specific examples for based on your own experiences.
And now AI, obviously, is such an important topic now. It's great to hear directly from you, where you focus on these topics, where you invest, how you choose entrepreneurs, how you are people first. These nuggets of wisdom are very valuable and I look forward to sharing them with the world.
Vijay Reddy: Thanks, Gopi. I think there's not too many times you can look back at your, like, technology career and say there's been such a big change.
And now it feels like 99, not just from the amount of activity, but the amount of tech disruption. And so for founders who are out there on the edge, I would say it's warm, water is warm. So jump in, but only if you are very, very passionate about something which you want to solve, right? There's enough Venture capital dollars out there, not just Mayfield, but there's a lot of investors who are [00:28:00] looking for good ideas and good teams.
If that's your calling, I think now is as good a time as any, and ping us if you fit the Mayfield model.
Gopi Rangan: The time is now, and now take a long term view. Thank you very much for that positive thought.
Vijay Reddy: Thanks, Gopi. Thanks a lot.
Gopi Rangan: Thank you for listening to The Sure Shot Entrepreneur. I hope you enjoyed listening to real-life stories about early believers supporting ambitious entrepreneurs.
Please subscribe to the podcast and post a review. Your comments will help other entrepreneurs find this podcast. I look forward to catching you at the next episode.