Ravi Gupta announces new company while staying at Sequoia, Pat Grady outlines why now is the best time to start an AI application-layer company

want to multiream, go to reream.com. And up next,

it's steward time.

It's steward time. We're stewarding. Uh we are joined by Sequoia Capitals, Pat Grady, and Robbie Gupta.

What's going on?

How are you guys doing? Great to see you.

How you see you guys?

Happy New Year. Uh we might need you guys to scoot back a little bit. Is that possible? We're just

We don't want half of you, Robbie. We want the whole Thank you so much. Uh uh massive start to the new year. Take us through the news today. Very exciting. Um but but we'll hear from you.

The news today about XAI is raising money. Is that the news that we're talking?

Yes,

that was yesterday actually.

Oh, okay. My

We've moved on. We already covered that. We're here to talk about

You guys are looking fantastic today. Just letting you guys know. Particularly good, I think,

dude. So, what happened today was, as you recall from my last TVPN appearance, um I want my children to be AI researchers so they can support the family. And yes,

it wasn't obvious they were going to get hired anywhere else. So, we're like, [ __ ] it. Let's I'm going to start my own company.

Epism, right? Epotism for the back.

Um those are the links that I will go to to, you know, make my children happy. No, what happened was today uh I announced that I am going to be starting a new company. I'm super pumped. And then also um getting to do it while still being part of Sequoia is pretty special and I think uh

my my brother Pat here, you know, was cool about letting that happen and uh we're excited.

Yeah. How did you two first meet?

Was it was it was it Abramson? Yeah, of course. I think we had a partner named Michael Abramson um who was involved with Instacart back in the day

and uh I think Michael was the original connection

and then Ruby we got to know RVY over a period of years and I mentioned in a tweet today that we always have this shadow list which is not long. It's like literally half a dozen people but you know the shadow list of the people we would hire if we ever got a chance was on that list for a few years and finally were lucky enough to get him back in 2019. Whenever someone asked me about my Sequoia interview process, John and Jordy, I said it was either five years or five minutes. I don't know. It was one of those seemed to happen quick at the end, but you know, I guess it took a long time.

Being gangsted by the Sequoia partnership. Uh I would love to see that list, but uh it's certainly uh key intellectual

property. Yeah. Key

stewarded. One of the things that you steered that's a steward's greatest responsibility.

How are you settling into the new role?

Uh it's been great. You know, we uh first off, Alfred and I have been working together for 15 years.

And uh he's been

success.

How do you two meet?

You guys are big on this question, huh?

I just like lore.

I didn't think we were going to answer it the first time. I thought they were

We could do, but I also like lore.

Well, I think the um I think most people know Alfred before Sequoia was the president and COO of Zapos, which was a play investment. And so we got to know each other a little bit through that, but I wasn't point on the investment. So it was just kind of seeing each other at Sequoia events or whatever.

And then he joined I want to say 2010 he joined Sequoia somewhere around that time.

And at that point I was mostly focused on growth investing. He was mostly focused on early investing. Um by 2015 or so he was co-aptaining the early business. I was co- captaining the growth business. And um we just had a great partnership for a very long time.

Yeah. How are you thinking about the divide of responsibilities now? Is there more nuance to it than just early and growth? Are there other is there someone who likes to do more personnel management, more recruiting or more marketing or whatever, more fundraising? Is there any division there?

So, once upon a time, Fred Luddy, who is the founder of Service Now, handed over the reigns to Frank Slutman. And I remember shortly thereafter, Frank Slutman had the line, uh, I'm the workhorse, he's the show pony. So Alfred and I have this wonderful relationship where the plan that we've agreed to certainly agreed to is I get to go on TVPN.

Okay.

He gets to show compliance.

You know, well, Alfred's welcome anytime, too. Uh it's fantastic. Well,

no, no, I'm kidding. I'm kidding. The the the honest answer is um our team, we have so many we have so many awesome people at Squa, not just on the investment team, but on all the operating teams. there isn't a lot of management that needs to happen. And so, um, over the last couple of months, I think one of the things we've realized is it's it's like very it's a very small amount of incremental load to, um, to do the the steward job. It's mostly just serving as a thought partner for other people in the organization who kind of already know what the right answer is. And so there's not there hasn't actually been that much incremental work, which has been a nice uh nice surprise. What is the what is unique about Sequoia's surface area of work to be done? I mean, I know there's the RAIA transition. Is there a public markets desk? Are there traders now? Like, how big is the organization outside of what people think of from like a classic VC partnership? What else are what else is going on at Sequoia?

Why you do that one?

Yeah, I mean, you know what's funny, guys? I think people would be surprised at how small Sequoia is. you know like the I mean we I think we barely have doubledigit general partners in the business right the entire firm is pretty small and so I would tell you this and I mean this sincerely like I think people tend to overestimate like surface area right of what we have in terms of like management scope

and realistically the most important thing is that the stewards are awesome investors and people that great people want to work with you know and so the thing I didn't say that's what you guys were I said that's like an important thing for you That's the ideal.

That's like the That's the standard. Now, now we'll talk about it. That's what they're supposed to be. No, but obviously like so truly like I would tell you the thing that's most important for these guys is to be great investors. Yeah.

It is to sponsor investments in the best companies and to continue that because like Sequoia's entire thing needs to be that we back the best companies in the world, right? And if we keep doing that, everything else kind of follows. If we don't, you're going to be the best manager ever.

Yeah. It's like it's like a a GP or or a steward that's not investing is like a CEO with bad product sense. It's like can they can they really be a truly incredible CEO if they don't understand the product?

Yeah. You lose your moral authority and we're too small of an organization. There's no there's no management job available at Sequoia. It's just too small of an organization.

Yeah. Uh Ravi, how are you thinking about uh the specific advantages of starting a company from your experience at Sequoa? There's obviously going to be endless industry plant allegations that you'll have to beat, but uh I'm sure you'll take those on head first.

Well, I guess one question would be

how much of the company do you build beforehand? How do you actually think about uh how you want to build this company? What are the axioms that you're following?

Yeah. Yeah. Yeah. Go ahead. Jordy, were you going to say something?

Well, yeah. I was going to say like what what qualifies as an opportunity that is worthy of of you know leaving an institution

because a lot of the ideas that you think are good somebody will come to you and be like I'm going to do it and you're like that's great I just write you the check and then you know you you go do the hard work but this is

well the pressure is immense like you basic you basically have to put on an absolute master class here right there's no excuses right you have you have to you have to achieve greatness

you don't think I can just yell at myself to grow faster and burn less not that's not how this is going to work.

That was one of my one of my favorite portfolio reviews is Revy Revy was unfurling this beautiful monologue about what was going on at some company. It was probably a five minute speech. It sounded great.

Yeah.

And then when you boil it all down, he basically said, "Grow faster, burn less." Five minutes of the thing happens at the ear. Okay. Um and you know, okay. So the real answer I guess the first thing is Jordy like you know this I'm not leaving. I'm changing my role, right? I'm going to go from like a general partner to partner on the boards, all that. But look, it is a high standard. This job is amazing. I love this job. I love our team and I love the founders I get to work with. I think the real thing that pulls it towards is not the specific idea. It's the specific moment we're in.

I do think that AI is changing everything.

I think that like I I wrote this thing maybe like a year ago, AI or die, which was just sort of like can you see the progress that's happening? And I really believe the world belongs to the fastmoving, reactive, responsive company right now that can get the best people in the world. And it doesn't matter if they're an incumbent or they're a startup. But I think that you like being able to move quickly and respond to that is the biggest advantage that you can have. And I found myself thinking about that all the time. And after writing that, I was like telling all these companies I work with and even companies I don't work with, like look, this is what I think it's about. This is what I think we got to be doing. And then I was like, man, I like really do feel passionate about this. I want to do that myself. And so the real thing that happened was I just wanted to be on the field again. I wanted to build a team. I wanted to go after this. And I think the thing that I uh found was that the fact that I could then do it while still maintaining my relationships with my partners, the fact that I can stay on some boards that made it so much more like something more I wanted to do. And then the other thing is, and I haven't shared this yet, and I'll share it with you guys when the time is right, but like I have a co-founder who's out of this world. Truly out of this world. Like way better than me. Someone that will make you guys bang the gong when you guys hear who it is.

So, are you going to be the show pony or or work?

I mean, I think it's going to be both.

It's pretty obvious the good looks of the operation. I mean, I'm surprised that you'd ask. Um, you know, but but put it this way, like when I told the kids like, "Maybe what do you think of dad starting a company?" They were somewhat like, "Well, I don't like you know you know what do you think the implications are do we have to move if this doesn't work that kind of

dude uh basically okay and I was like how would you feel if this person was the co-founder and then my kids were like well we actually have $750 of accumulated birthday money maybe we could lead your

get him on the sequoia list for future

they did not call me but uh Andrew's tweet was an alltimer today all time

so good how how are thinking about uh timelines with the new company. It feels like you can build products so much faster today than maybe you could two years ago that there's maybe more of a a value in in staying uh really really quiet, actually staying in stealth and and making some more uh progress before you come out the gates and then invite a bunch of people to try to come compete with you.

Yeah, I think listen, we're going to talk when we have something to say, right? We got to go and build this thing. I do think to your point that you know the Sierra guys are very good about this and they have this language of like in AI world a year is a month right or you know whatever whichever way make it faster but basically like a month is a year you know uh you know a week is a month and a day is a week and the idea is like you got to go fast you need to be super responsive you guys know I'm big basketball guy coach K has this great quote you know I'm not a world class predictor but I am a world class reactor like that's what we got to be and you got to be fast and truth seeeking And I think the other thing that's nice about being, you know, with Sequoia is like we get to see the best companies in the world, you know, and we get to learn from them and see the pace that the best ones go. And, you know, Pat, you can talk about that, but I I think that's certainly something that is helpful.

Yeah. How do you think about uh the particular AI opportunity right now? I feel like there's there's a little bit of a meme that the best time to start a a AI company is probably 10 years ago as a lab and wait around and then have this breakout moment and the second best time was probably like the day Chat GBT launched and everyone sort of realized that this was real but we're now two years into this a lot of you know the the the default categories the market maps exist. How do you think about just finding opportunity? Is it unlocked by just last year's advances or do you think there's new categories that were not on the table two years ago that are now on the table? And I mean, this is your thinking, but also for anyone that's watching uh who might want to start a company in 2026.

I I kind of think that the premise that those other people have is flawed,

right? And I'll tell you for real, like the biggest thing that makes a difference in whether a company succeeds or not is the people that are building it,

right? And the I believe that we have the most fierce competition ever, you know, right now. And it's only going to get more fierce, right? And I think that you're gonna be fighting every day. And there's no like actually Sarah Pat's wife wrote this awesome post yesterday about like uh relevance decay, right? And how much more quickly it's happening. And basically like you have to fight for your relevance every day in the venture world.

That is even more true at the company. Like it doesn't matter what you did yesterday. Look at think about recently the the goodness that's happened from Google, right? The best thing that happened to them is the competition that they have from OpenAI cuz all of a sudden it lit them up and then they had to go and rein all this stuff and re that mojo back. So even like the deepest moes in the world, you got to go fight for them every day. So the real answer you know John is it all we want to do is be in a space that we are willing to fight for every day and apply the latest technology every day and have AI native people working on it every day so that every day we get we fight again and earn our spot and I I in my heart I believe that more than anything which is that like there's nothing to protect right now even at Sequoia there's nothing to protect the only thing is what we go do tomorrow and that'll be true at the company too. So, I think that like honestly I think when people are like, "Oh, the best time was before." I think that's like low agency thinking and I'm like, I don't have any time for that. Like you got to go out and get it.

Yeah.

I can make a specific case for why now's the time, too.

Please.

So, the I think if you want to start a foundation model lab, yeah, you're probably better off to have already done that. You know, we kind of already seen that play out.

Yeah.

One of our core hypotheses in the world of AI is that the values at the application layer. I think even for the foundation model labs, we're starting to see that that's true. Um, I would argue that now might actually be the single best time to start an application layer company. Not because everything is whites space, but because it feels like the playbook is finally starting to emerge. And specifically, if you think about like the three major inflection points that we've seen in the world of AI. I think the first was November 2022 when uh chat GBT came out and the world was exposed to pre-training.

I think the second was fall of 2024 when 01 came out and the world was exposed to reasoning. Mhm.

I think the next one was actually the last couple of weeks where you see cloud code with Opus 45

and all of a sudden we see what long horizon agents can do.

Yeah.

And it's hard to like we're venture capitalists. We're not in a position to try to define AGI. But from a functional standpoint, if you have a piece of technology that can just like keep going and overcome obstacles and use resources and eventually think its way to the outcome that you wanted. I don't know, man. seems pretty close to AGI for me. And I think if you

we have we have goalpost physical goalposts here that we'll continue.

Yeah.

Yeah. Exactly. And so and so anyway, I think now that that playbook has kind like not quite been written, but you can kind of see it, you I think I think the ability for these application layer companies to go from a chatbot that in best case scenario people use four or five times a day to an agent that is always on running six instances in parallel 247 doing the work of six human beings like there's just a dramatically different impact that application layer companies can have today using the most recent playbook versus what they could have done two or three years ago. So I think for that reason it's a great time to start a company real quick. Yeah. Yeah.

Dude, I I think that I so agree with that. I think also if you just kind of think about this idea of like a 10x engineer that has become so popular in Silicon Valley. I think Steve Jobs talked about it. Whatever.

I think that the leverage available to great people is it's never been higher and it's only going to get higher. Right? All these things you guys see about the value of agency and it's like all about like how much do you want to get your idea into the world. I think that I don't know how to do the math perfectly of like is it a hundred times person or a thousand times person but I think that the leverage available to the best people is insane. That's a huge aspect of the new thing that I want to build too and I want to build with our co-founder because this idea that I believe in of like people are power laws, right? People follow power laws. I think that's true and it's only going to get more defined. And so what Pat's saying about like you know the leverage you can give to somebody with today's technology has never ever ever been higher.

Do you guys have any type of internal philosophy around it feels harder than ever to predict like 2 years out? Like I think in 2020 you had a good sense of like what the world and what software was going to look like in 2022 and it feels much harder to do that today. Do you guys have any internal philosophy, you know, similar to Bezos thinking around people are going to want things like cheaper and faster and I'm just going to focus on that. The world's going to change, but people will still want things that are cheaper and faster. Do you guys apply any thinking like that around when you're looking at investments Pat maybe at the at the late stage or Ravi with with with your new company around like okay a lot of things are going to change potentially you know we can imagine a century of progress in you know 20 years or 10 years or eventually yeah two hours uh who knows um so how do you build something that's going to be durable and and again like relevant uh as you get increased rate of change

what's your thing on inevitable Yeah. Yeah. I think you I think there's a few things. So, yes, I think and I think Pat and I agree on a bunch of this uh too, but I think one you want things that are early and inevitable, right? Like I think that there are things that when you see them, they're like, "Well, that definitely is going to happen." And just to like go back in time, I think, you know, when I joined Instacart in 2015, this idea of like, are people really going to shop for their groceries in 100 years the way they shoed for them 80 years ago? Like it certainly doesn't seem like we will do that. it it you know this is inevitable that it will go and it be different. I think that to your point it is harder to figure out what things are inevitable right now because the world is changing super fast. So I think one you got to be like more limited in that right of you know what are the what are the industries that will be durable. I think the there are two things I would tell you to answer your question though. One, I think network effects are still potentially even more powerful than they've ever been, which is I think if you get a company with a network effect and you execute with excellence and speed, I think you can keep going and build it even up more. And I, you know, never want to predict what the AIs are not going to be able to do, but I think it would be hard to say press this button cloud code and create a network effect of highly fragmented buyers and sellers. And so I do think network effects can endure. And then I think the biggest thing is the people which is you need the people that you will trust to go and face these realities. And it's the same way honestly like I know I got called on Twitter today but I'm going to I'm going to go with this dude. It's the same way with your kids which is you want to prep them for every possible situation that's ever going to happen in their life.

But the reality is you can't. And all you can do is make sure that they're actually good at reacting to tough things as a general matter. And that's the same thing I think that

in a value system.

Yeah.

Yeah. What's the changes we're going to see? What you want is to trust that the people you have are good at responding to those things with truth seeeking, with agency, with aggression. Uh and I know on that dimension, I know Pat and I agree.

There's also there's a Jamie Diamond line which is it's easier for me to predict the next 10 years than it is for me to predict the next quarter.

And part of what goes into that is the forces that shape the next 10 years are kind of so big. and so obvious that you can actually make in some ways higher fidelity predictions over a longer period of time. And I think similarly with our business, if you look at um like for example, can you imagine a version of the future in which doctors aren't using all of the world's medical knowledge on an application in their pocket to make better decisions at the point of care? Of course not. Like of course that's going to be a thing. And you know, like there happens to be a company, Open Evidence, that is doing that today for a decent chunk of the doctors in the country. But like obviously that's going to be a thing. Or by 2050, are people going to drive cars in 2050? I have a hard time believing they will. Like it's going to be the way people ride horses today. Like you might do it as a novelty, but it's not going to be the way you get, you know, to and from point A and point B. And so there are some things where you just like know that it's inevitable and it's a question of okay, well, what's the path from here to there and who is it that blazes that path? And that's where it can be tricky. And I think that's where RVy's commentary on the people really come in because one trap that it's easy to fall into is theorizing on market structure and different technical approaches. And the reality is if you can figure out the best people, they're going to figure all that stuff out. Like they'll they'll figure out the right technical approach and the right business model and all that stuff. Like if you just figure out the right people, they'll probably find the path. I think one one challenge right now uh for founders that are trying to navigate the world is like you know 10 years ago if you were looking and trying to understand the dialogue and different narratives and conversations happening in tech it would be like 99% kind of like market analysis product analysis 1% science fiction fanfiction and now the the narrative and and the conversation is like 50/50. It's like some like kind of like tangible analysis of what's actually happening and how markets are evolving and then 50% like sci-fi fanfiction of people writing like it's possible that in two months a galaxy like you'll be able to buy a galaxy. What does this mean for so like you kind of need to you you kind of need to like drown that out and like pay attention like that goes back like just talk to customers what do they what do they need today for sure

react to the the game on the field. Do you guys think that uh we were debating earlier it feels like AI is is doesn't have its Steve Jobs from a messaging standpoint? Like I would love an hour an earnest, you know, conversation of an hour from a from Steve Jobs or or a Steve Jobs like figure talking about because I feel like society broadly underrates AI. They see the slop videos and they see the idea that their electrical bill might be going up and they're like, I hate this. or they see the slop and they they hear they might lose their job in a couple years or they can't find a job and they're like I hate this and it feels like we don't have anybody right now that can go on that can go like when when you know as as great as the Elons and the Sams and and the Darios and people like this are as operators and builders and visionaries when they go on podcasts and those videos go out the comment section are all like please no stop turn it off you know and so it feels like we're missing we're missing a

maybe it' Impossible. I don't know. Yeah. What do you think?

Oh, uh I I I agree with you. I think we we are missing that person. I think Pat and I were both in the meeting yesterday though that this kind of reminds me of and I will share it on his behalf. But like Alfred shared a story the other day, yesterday about his dad. Okay. And I think it is pretty instructive in terms of this thing. So Alfred's dad, I don't think this is going to surprise you, was a pretty brilliant guy, right? Um, but what his job was was to be a human human calculator. Okay? Like literally like sit in a row like do math, like you know, be a cell in Excel, right? And just be a human calculator.

And at some point, what technology allowed was for him to not have that job anymore, right? And someone with this giant brain and this ability to produce and help society could all of a sudden not do math and put numbers into one cell anymore. And he could have way more impact and build an accounting system for the company that he worked with. And all of a sudden all those other people who have huge brains, right, could go do that because of technology. And he made this point which is like look technology is neither good nor bad. It's what we do with it, right? The same technology that powers a nuclear bomb can power a nuclear power plant, right? And so the idea it's like what do we do with it? What do we do with it? What do we do with it? I do think that a lot of what people are doing when they go and talk about the technology is they talk about all the things that might happen. I would rather us talk about the fact of it's going to get more and more and more powerful and it's more and more incumbent on all of us to make sure we use it for something that's great.

And I actually genuinely believe that like I I' I've written about this before, but like you know if I could give my kids one thing, it wouldn't be kindness, it wouldn't be curiosity, it'd be agency. And this is that same point which is like listen the technology is going to be what all of us make of it you know and I think Alfred's point sat with me which is like he talks about his dad being able to do more with his life so that his son could do more with his life and a lot of that's due to technology like nothing better than that.

Yeah. Well that's a great place to wrap it. Thank you so much for taking the time to come chat with us and congratulations on the the the move and everything. We're very

super excited for the new company and evolution of the role. buried the lead, but we got to have you back on as soon as we know more about what you're building, and we would love to do it here live. Uh, but congratulations to you both on all the progress. Hope uh 2026 is a fantastic year, and

happy new year to the whole team.

Happy New Year to the whole team.

Thank you guys.

Cheers.

Uh, he mentioned Jamie Diamond. There's some breaking news about Jamie Diamond. JP Morgan is taking over the Apple Card program from Goldman Sachs. That's very exciting. Uh, also let me tell you about Turopuffer serverless vector and full text search built from first principles