The GTMnow Podcast

VC: Investing at Inception in the Age of AI Agents | Ed Sim (Founder & GP, Boldstart)

• GTMnow

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 49:46

Ed Sim has been a VC for 30 years. He's backed companies like Clay, Front, BigID, and Snyk. He writes What's Hot in Enterprise IT every single Saturday, 489 weeks in a row. And right now, he says this is the most exciting and terrifying moment he's ever seen in his career.

In this episode, Max and Ed break down what's actually happening inside startups and boards right now, why the old playbooks are dead, and what separates the companies that will survive this AI shift from the ones quietly getting killed by it.


Discussed in this episode

  • Why engineering is no longer your bottleneck (and what is)

  • The 5 P's Ed uses to evaluate every inception-stage investment

  • The autonomous enterprise thesis and what it means for how companies are built

  • Why AI-native leadership is now a survival reqxtuirement, not a nice to have

  • The full Clay story: $600K to $100M ARR, how they stayed lean, and what actually unlocked growth

  • The 3 CH's framework for being a great board partner to founders

  • Why the best founders today are inside the AI jet stream, not chasing it

  • What every board meeting sounds like right now

Episode highlights0:00 Intro & 

1:05 Episode Preview: Ed Sim & Key Takeaways

3:10 The Jet Stream Analogy: Two Types of Companies

5:43 How GTM Operators Should Evaluate Companies Like Angel Investors

7:20 The Collapsing of Moats & AI-Native Business Opportunities

10:00 Rebuilding Industries vs. Selling Software to Them

15:00 Why Old GTM Playbooks Are Dead

17:46 Ed Sim's Background: From Cutco to 30 Years in VC

21:43 The Five P's of Inception Investing

23:04 How to Evaluate Potential & TAM in a Fast-Changing Market

25:40 Staying Ahead of the Jet Stream as a Founder

26:32 The Autonomous Enterprise Thesis

28:44 Agent of the Week: How Companies Should Adopt AI Agents

29:10 How Agents Are Changing Engineering Bottlenecks

31:15 What Incumbents Must Do to Survive the AI Wave

32:53 Intercom, Snowflake & How Legacy Companies Are Adapting

36:43 The Clay Story: How They Found Their Footing

38:33 The Three C's of Working With Founders (Cheer, Challenge, Chill)

40:07 Clay's Growth Trajectory: $600K to $100M+ ARR

41:10 Clay's Agency GTM Model & Community Moat

43:50 Ed's Fund Model: $500K to $15M Checks at Inception

46:57 What's Hot in Enterprise IT & Venture Right Now

48:03 Closing Remarks


Key takeaways1. Engineering is no longer your bottleneck. Your people are. Code is shipping faster than your sales, marketing, and customers can absorb it. The constraint has flipped completely and most companies haven't noticed yet.

2. Painkillers beat vitamins every time. The only startups worth backing at inception are solving a hair-on-fire problem someone desperately needs fixed, not a nice-to-have they can live without.

3. The 3 CH's of being a great board partner. Know when to Cheer (when founders are getting beaten up), when to Challenge (when they feel invincible), and when to Chill (when they just need breathing room to figure it out). Elliot used all three with Clay to perfection.

4. If your CEO came from sales, you are in trouble. Surviving this AI shift requires product-driven, agent-native leadership at the top. The companies that adapted, Snowflake, Intercom, Atlassian, all changed leadership first.

5. The best founders are inside the AI jet stream, not chasing it. The question Ed asks every founder today: are you struggling to keep up, or are you the one constantly shipping and adapting faster than anyone can copy you?

<

The GTMnow Podcast
The GTMnow Podcast is a weekly podcast featuring interviews with the top 1% GTM executives, VCs, and founders. Conversations reveal the unshared details behind how they have grown companies, and the go-to-market strategies responsible for shaping that growth.

Visit gtmnow.com for more episodes and other interesting content. 

SPEAKER_03

I'm gonna believe it. There's two kinds of ways to make money in this world. It's basically you go big and you see other people going bigger and bigger and bigger, and you go niche specialized or you go home.

SPEAKER_02

Today we're with Ed Tim, founder and general partner at Bold Star. Ed also writes a weekly newsletter called What's Hot in Enterprise IT. 489 weeks in a row. Right?

SPEAKER_03

Yeah, consistency, you know, every Saturday morning I pump it out and it helps me synthesize kind of what's in my brain and what we're seeing.

SPEAKER_02

For a founder that they're like, I'm gonna go all in on this thing, and little do they know Claude is doing the release next week that's gonna make it obsolete, and three other companies just raised $500 million that are winning in that space, right?

SPEAKER_03

I know that most things can go wrong when you invest, and the earlier you go, the more things that can go wrong. Let's say 9,999 things can go wrong. If one thing goes right, what does it take? And how big can it be?

SPEAKER_02

This is our bonus podcast on GTM Now, and we are here with Ed Sim today. But first, you've got Paul Irving, my general partner. Congratulations. I think first episode with the new promotion. How's it feel?

SPEAKER_00

Thank you. Thank you, sir. Feels great. Got a lot of wonderful notes back from our founders, investors, people we work closely with. I don't know if this is a good or a bad thing, but 24 hours later, somebody on a Zoom meeting said congratulations to me, and I totally forgot what they were referencing. So back to work and back to building as it always is, but it was a fun one to be able to announce. And obviously, we talk about it all the time, but never been more bullish in uh what we're building together at GTM Fund.

SPEAKER_02

I think when you don't know what the congrats is for, it just means that there's so many good things happening all at the same time. So that's a plus. Uh, we got Ed on the show today. I thought it was a phenomenal episode. Uh, certainly one of our better ones. Ed's a legend, been in quite a few investments. We talk about the five Ps and three CHs, and this is one of those episodes where I actually thought that I learned a lot personally and like some things that we're even going to change in our fund, which I really enjoyed. It's another one of those episodes where I also said to myself, every GTM leader, everybody in tech even should listen to these episodes so they could understand how investors are thinking about markets. Like, if I was a CRO right now or a VP of demand gen, and I was like, Oh, I want to go pick my head up and find my next role in my career. You should want to go where a lot of very savvy VCs have been thinking about where the ball is going. Like your career is being like a VC, like your time is equity. You're investing with your time. So you should go to the best possible company you go to where your time is going to be worth the most amount of money in the form of your equity. And you should start maybe in a space that is exploding. And he kind of goes into this jet stream analogy, which I thought was pretty good around that. So, yeah, I mean, what what were some of your key insights out of the episode? I think you know, the jet stream one was a specifically uh resonated with you quite a bit, right?

SPEAKER_00

It did. And I'll jump into the jet stream analogy because I thought it was a great way to look at, especially companies today in an AI native world. But to your point on operators, we've got a bunch of incredible, you know, GTM executives and operators in our network. And the ones that repeatedly choose the best companies time and time again are people that look at it from the framework which you mentioned, which is almost an angel investment framework. And you still do the diligence you should do as an operator in any function. If it's sales, you want to look at pipeline, you want to look at close rates, you want to look at the RevOps stack, you want to talk to some of the ICs within the organization to get a feel for, hey, can I run this go-to-market org and and you know, can I meaningfully up level what they're executing here? But then you almost want to step back and look at it as if you're an angel investor, which I think brings into a number of other layers that are not just go-to-market oriented and can I sell this product? And do I think it's a cool customer experience? And do customers like it? But what does this market look like? How competitive is it? What have exit opportunities look like? Who are the investors? And then what does the next round look like if we're successful? And when does it happen? The same exercise we would go through as an investor would be such a valuable one for operators or even higher-level ICs to do the same on their side when they're they're deciding where to go next in their careers. This is probably a a question I would add to my analysis, which was a framework that Ed mentioned and agreed this was a fantastic episode was the Jetstream analogy, where there's two types of companies. There's company A, you're building a deep technical product, maybe a non-consensus thesis. In every startup, you have to move quickly, the speed of light these days, and you have to build, but there is almost some aspect of time for you to build that product and figure it out. And then there's other ones where you know you are dropping straight into the jet stream, which is it is consensus. The foundation model companies or frontiel model companies are building in that direction. You know that the demand is there, you know that if you capture it, you can hockey stick in the right direction. But the challenge then becomes how do you keep up with the jet stream? How do you stay on the front edge? And so these are the replets and lovables and cursors of the world, macores of the world. The best founders and teams know how to stay ahead of the jet stream, but you should go into it, whether you're joining a company or your investor looking at a company, understanding which you are. I don't think there's a right or wrong answer. Like some people will say, never put yourself in front of the jet stream. There's a lot of iconic, incredible companies that are being built right now that are putting themselves in front of the jet stream. The only mistake I believe you can make with that is not knowing whether which company you are. You have to know if you are putting yourself in front of the jet stream and you have to operate and execute everything across your business along that principle. And if you're not, you know, you're gonna build your company, the timeline in which you do it, the people that you hire is gonna look a little bit different too.

SPEAKER_02

Yeah. You know, I had a journeyman CRO say to me maybe two weeks ago now, you know, I'm on my whatever, 10th startup, third CRO, head of sales role, and you just have like just haven't got my big payday yet. And I've watched a lot of friends get paid and you know, whatever over the years, and it's been frustrating. And I actually think that they're pretty talented. And you know, the funny thing they said to me after that was like, it's a bummer because if I had just like gone and taken an IC role at Anthropic three years ago, I probably would make more money in my entire career than like everything else that I've done so far. And it's just, yes, it takes a little bit of luck. Like if you had had the foresight that that would be the company then, sure, like then you know, you've earned it. But I do think you can get more, I'd say methodical about how you choose great companies and great spaces to go work in. I think at the end of the day, the old axe expression from Abraham Lincoln, right? Or I think everybody misassociates that, Mark Twain here, who knows? But if you give me six hours to chop down a tree, you know, I'll spend five hours sharpening the axe and one hour chop down the tree. I think when you're sizing up what you should do in your career as an executive, you should be thinking like a BC. And so, yeah, you should be partnering with those founders who are are in that jet stream. I think, you know, uh one of the things we're seeing right now also is this collapsing of motes. I mean, we talked a lot about that. You know, we we talked about what still counts when it comes to moats. And he said that uh finding a domain complexity that cannot be collapsed into a simple skill. That dovetails into some other topics you and I have been talking about for a while now, which is, you know, where are the big opportunities? And you know, instead of just being in this world where I think for the you know 2010s, you would just see like the 10th marketing automation company, HubSpot competitor. Okay, like well, what do you do better? It's like, oh, or like, you know, cheaper, faster. I like what what like what are you gonna do? Okay, we're vertical vertical focused. Okay, great, but like how are you vertical focused? Well, we do this like we have a couple special API integrations with you know players in the prop tech space. So like we're just easier to use and connect to a lot of the other stuff that they use. And it's like, oh okay. And now you're seeing what AI can really impact is like instead of going into a vertical and industry and supporting them with software, you can almost like rebuild the industry with AI, like it's complete reimagining and rebuilding. And he touches on that, you know, with us in the episode. But we've been talking about that for a while, and then the YC26 batch just came out. And if you look at that batch and you look at the themes that are in there, you're starting to see kind of the emergence of that. How do we go into healthcare? How do we go into petrochemical or nuclear space and not just build software for these industries for the companies that exist in them, but like create the industry, rebuild the industry?

SPEAKER_00

It's a completely new change of perspective, which I think is an exciting one to a degree that I don't know if we've ever seen yet in technology, where and even if you look at software historically over the last 10 years, so many of the even very successful companies built for other software companies' problems. And you would see in a variety of different verticals, really strong products, companies, some companies who have gone public are almost there, but they are for the most part selling to other technology companies and they're very successful at it. But crossing the chasm into other industries is difficult. And the ambition of founders we're seeing today in the early stage, and you see a YC is always a perfect microcosm of this, is you get some of the youngest, smartest, most ambitious people. What are they thinking about and what do they believe they can build, you know, and and what does their future look like? And you're starting to get things that look less and less like a traditional software company and more and more like AI native business opportunities, if you want to call them that. So instead of, hey, here's software to or help with the discovery of mining and minerals, we can do an end-to-end uranium discovery platform. And we're using technology, we built proprietary in-house models to be able to execute on that vision. But at the end of the day, we we look like a technology native, AI native integrated version of that business where we're not just selling technology to those end users within that vertical, but we're able to execute for your end customers and take a huge swath of that workflow and value chain underneath our umbrella.

SPEAKER_02

It is the work. It is in the past, you would see a company that's like, we're building voice and dialer and dashboarding for call centers. Okay, so buyer software, we have a thousand call center reps. Reps sits in the seat, uses your software, hits the phones. Now it's AI call center worker. The agent calls, manages the entire thing, right? Like end-to-end. There's AI mortgage brokers, like it end-to-end essentially. And I do think there will certainly be some of these things that are like, well, as long as there's a human on the other side, they'll, you know, a human may always want to have that conversation. And there are other things that we've said for a long time. Why does a human need to be involved in this? Like, why am I spending so much money on human here? And a good example of that might be legal docs for certain things, right? I think there are certain things you definitely want a lawyer for, and you want to be able to speak to person. You want to be able to understand kind of the nuances of the situation. And there are other things where it was like, I have to pay you like $1,500 just to give me this like boilerplate document with a bunch of my stuff filled in on it that you do for everybody else all the time. Like this is standard practice. Let me just get that. But I think that's what's happening, you know, across the board and in like in kind of the simplest terms. And we're seeing it in a lot of different industries, and it's you know, fascinating ones. You just need mining, space, robotics. So what else do you see? What else stuck out to you from kind of the YC batch and you know, from a market standpoint, you know, how how are you seeing that play out across you know kind of the rest of the ecosystem?

SPEAKER_00

Yeah, there's a couple interesting there was maybe a few less surprising demographics from the YC batch. Like, you know, a year ago, maybe the speed in which we got here is fascinating, but a year ago it was sort of co-pilot for everything 18 months ago, and now it's just fully agentic doing the workflows end to end. Not surprising, but just interesting to see the level of penetration. The interesting stuff I wasn't necessarily expecting going into it was healthcare being the biggest vertical, 22 companies total. Obviously, one of the largest areas of spend in the US economy and global economy, a heavily regulated industry. So I think part of the thesis there is if you can build the dollars in the budget are there, the impact you can have on people, communities, you know, the world at large, if you build a, you know, agentic healthcare company that allows, you know, emergency rooms to operate more efficiency or the doctors or nurses within them, that's really impactful. But it's also defensible. There's data, there's workflow, there's compliance and regulatory guardrails in place. So not surprised to see more companies, but the extent of them was interesting. Physical AI, which you mentioned, you know, robotics, mining, agriculture, warehousing. And then the last sort of two things that stood out to me was rethinking the internet for an agent economy. So many old platforms were about human access, human-to-human payments, but facilitated through software or through the web, software monitoring for human behavior and activity. But when so many of these functions are done by agents, you now need agentic identity verification. You need agentic security, agent-to-agent payments and billing, agent monitoring. If you start to look at just how much of the web is going to change and commerce that exists across technology platforms is going to change and it's going to be fully agentic, then you would rebuild so many primitives. There's a bunch of infrastructure companies there. And then the last bucket was just solo founders. I think for a while people have talked about the one person or two-person billion dollar company. There was 27, I believe, solo founders in the YC batch, one of the highest in a long time. And I think you're starting to see that principle play out. People are going for it. You know, they're going to build their teams to execute, but there seems to be less, you know, deference to I need to find a co-founder than there was historically.

SPEAKER_02

That's a perfect segue into the show. Today's guest is Ed Sim, as we mentioned. Ed's been investing for 30 years and he's been doing a lot of cybersecurity and enterprise investments. He's in sneak, big ID, customer clay, early in his career, live person and go to meeting. Without any further delay, let's get into the show with Ed Sim, founder and general partner at Bold Start. All right, welcome back to another episode of the VC special edition podcast we do here on GTM Now. Today I'm joined by Ed Sim, founder and GP at Bold Start. Ed, how you doing? Good. Thanks for having me. Yeah, definitely. Uh, somebody I followed for a very long time, must follow on Twitter. Common theme here with uh a bunch of these episodes, but at Ed Sim, nice and easy for everyone. And also writes a weekly newsletter called What's Hot in Enterprise IT. 489 weeks in a row. Is that right?

SPEAKER_03

Yeah, consistency. You know, every Saturday morning I pump it out and it uh helps me synthesize kind of what's in my brain and what we're seeing. And frankly, I think we're seeing way too much right now, Max. What about you? Uh just yeah, something we can talk about later.

SPEAKER_02

Good writing leads to clear thinking. So I appreciate you sharing it with the world, you know, first and foremost. But also, yeah, it's something I'd say uh not as in depth as you, but something I try to do every week is Yeah, I love your newsletter, by the way.

SPEAKER_03

You have great comments and great interviewers, uh interviewees, and uh lots of good aspects, I think, on GTM, because I think as you point out, and as I think about it too, the playbooks of yesterday are scrapped. Yeah. And right now, I think you just have to learn from what everyone else is doing, and also I think you have to have your own twist.

SPEAKER_02

Definitely. It's you know, kind of like everything else in AI these days, changing week to week. And it's certainly hard to keep up. The beautiful thing, and you've probably seen this yourself in doing this for 16 years plus, but you know, when you work with founders, you work with kind of the best and brightest people, and they're all tinkers. And so, you know, we're able to bring kind of our tried and true playbooks and even some new things to the table with them. And then it's amazing how they take it and run with it. And then we almost like end up learning more from them, and then we're able to bring that back to the rest of the portfolio. And so it's kind of this beautiful flywheel. Do you see kind of the same way with you guys? And even if it's not GTM, it's just, you know, other things that you pick up along the way.

SPEAKER_03

Oh, 1000%. And part of just throwing what I see and what we talk about, you know, behind closed doors. I won't share names, you know, in board meetings or with founders or in pitches, is to kind of just draw that out to maybe start that discussion. Cause I none of us know everything, right? And it just it's important to hear what else people are talking about. And, you know, the people that I'm sure are subscribed to your newsletter and mine are are craftsmen, right? They want to spend time trying to get better at what they do, which is why I send it out Saturday morning. If you're interested, it's already a signal that that you're reading it, that you care deeply about kind of what you're doing and you love what you're doing.

SPEAKER_02

Our our thoughts are similar. Ours goes out Friday, and it's like this is weekend reading material. Like if you're the best of the best, you're gonna sit down sometime in the next 48 hours and dig into this and you know, journal on it or bring it to your team or share it around. And you know, those are the people you want to surround yourself with, especially as you know, an up-and-comer and in the industry. And I think everybody's a student of the game at all times. Like you, you know, and and especially in this like kind of AI wave or era, whatever you want to call it that we're in, you have to be a student of the game. I mean, things are just changing so fast, right? It's there's I wouldn't say there's any gurus or anybody who's super knowledgeable in this area that's just like you know, that knows it all because then the next week the next thing comes out and it's like, uh oh, I gotta go learn something new.

SPEAKER_03

Absolutely.

SPEAKER_02

Yeah. So we'll start with a little bit of your background. So uh cut co sales rep, love that. Into investment banking into VC. Give us a little backstory.

SPEAKER_03

Yeah, look, I my parents are immigrants from Korea and I came with kind of just a trunk and my brother and grew up in Maryland and was always scrapping by. My parents taught the value of education. Being in Baltimore, one of the things that I found out was that lacrosse was a very popular sport. So I started playing that when I was very young, and it helped me get recruited into college, you know, along with having good grades. So I ended up at Harvard and played on the lacrosse team all four years. We're number three in the country, by the way, my freshman year. They're they're actually number four this year in the country right now. They beat Q. So that's that's kind of cool. But but the point is when I was there, I got exposed to a lot more than just kind of my small world in Baltimore. And one of the coolest things was having career days during, you know, every year. And a friend of mine's brother came up and showed up and talked about this thing called venture capital and uh explained to us what it was. And the idea was that hey, there's a founder with a dream and an idea, and they can't get capital and you believe in them, you give them capital. And you know, the interesting part was that you could either be a founder yourself, but that's a lot of risk, or you could spread it across kind of you know many different companies and perhaps you know, get involved where where you could, you know, in terms of helping them out, but at the same time, it's you're not tied to one thing forever and ever. And that that kind of really appealed to me. And so that's kind of how I learned about VC, uh, was back in the early 90s. And and I've been a VC for 30 years, believe it or not. So happy to dive in more about how I ended up there.

SPEAKER_02

Yeah, so you're doing investment banking previous to actually it was it was investment management.

SPEAKER_03

We were I was on the I was on the derivative side of JP Morgan, so I just basically got a job in the city. We're managing, you know, 20 billion dollars uh using futures and options. And we created, well, I wasn't the one creating it, but I was the one building spreadsheets uh all day long, and basically using this risk premium model to allocate assets between different countries based on kind of the difference in returns and stocks and equity and and fixed income and things like that. And what you do is when you're in spreadsheets all day, you learn how to code. So I learned Visual Basic. And then I started downloading a mosaic browser in '95, which was the precursor Netscape. And I was like, wow, okay, I think I can get into this venture capital thing because this internet thing is gonna be really big. And that was '90, '95, '96. And I ended up getting a job in VC in early '96 at a small firm in New York that was backed by the New York City government to build New York into a high-tech sector, you know, kind of that's what they called it back then. In 98, I started a fund with a guy named Bob Lesson, who was vice chairman of Smith Barney. And the idea then was to bring a Silicon Valley approach to investing to New York, which meant focusing on product, market opportunity, and the future versus looking at spreadsheets. And that was kind of the big difference. And so from then on, I've been doing the same thing since '96, so 30 years now.

SPEAKER_02

Yeah. You know, I saw from your background, or your investor in live person and go to meeting, which was kind of like a couple errors ago. And then, you know, front big ID customer, sneak, and a couple others that were kind of in let's call 2010 to 2020. And now you're investing obviously in net new companies now. Like, what have you seen? What are the common threads that are like tried and trued that it lasts that entire, you know, 30 years of investing? And obviously, what's changed a lot?

SPEAKER_03

When I first started investing, people I said that was about the people. And I, you know, obviously when people say that, you don't know what it means. Is it their resume? Is it kind of what they look like? I mean, you just don't know what it means. And I think for me, it's always starts with the people, particularly because when I've invested, it's always been in ideas. And they had different names back then. It was series A, then it became Seed, then there was pre-seed. I call it Inception, and we can talk about that later. But it's basically partnering with two or one technical founder who has a unique insight that is usually born out of pain. And because it's born out of pain and because they're technical, which they usually try to do is automate that problem away. And it's a type of founder that's on a mission. They can't stop thinking about it when they wake up in the middle of the night, when they take a shower, when they're out, you know, kind of having dinner sometimes. So their kind of brains kind of moving. And so I think for me, it starts with the people. And then I've kind of bulleted it down. Like I'm a frameworks kind of guy, and I came up with something called the five Ps for inception investing. And, you know, one is the people, two, it's the product. Kind of what is that unique technical insight? What is the magic that you're gonna bring into a customer's life that you know is gonna basically make them 10 times better? With your product or solution than without it. And of course, you've got to address a burning pain, right? You know, people want painkillers, not vitamins. And so are you identifying a hair and fire problem that is going to endure? And, you know, the bottom line, Max, is that these things don't go in straight lines. So, like we can talk about clay later, but you know, those guys wander around the desert for a little bit, just building product, ranking away. But you have to have passion because you're going to hit hard times and what keeps you going? And it's going to be really freaking hard. And so when you're on a mission to kind of build something and you don't want to quit and you want to keep going, I think that's super, super important. And the last part, I think, which is a big question that we'll probably hit on later, is just from an investment perspective, I call it the art of the possible. That's the fifth P kind of what's the potential. I know that most things can go wrong when you invest. And the earlier you go, the more things that can go wrong. Let's say 9,999 things can go wrong. But if one thing goes right, what does it take? And how big can it be if that is right and you're correct? Right. And so those are things that to kind of look at and they vary over time, but I think that's kind of how I think about it.

SPEAKER_02

We're gonna have some fun with clay in a little bit, but I want to stick to that fifth P there. So in a market where things are changing, you know, weekend to weekend, how do you size up potential, TAM, MOATs, all those types of things that kind of go into potential? For example, you know, you had a tweet, I think it was uh still seeing too many startups that will simply be a skill in days, weeks, or months. Make sure yours can't be reduced to one. That feels like it fits right in that fifth P because you know, for a founder that, you know, is working on something, they're passionate about it, they're like, I'm gonna go all in on this thing. And little do they know that Claude is doing a release next week that's gonna make it obsolete. And three other companies just raised $500 million, you know, across three rounds in one year that are winning in that space, right? So when you're kind of underwriting or looking at a company and they've got the passion, they've got all the other P's, but you know, you look at this potential piece, how fast we're moving right now in the A era factor into your judgment of, you know, that business and that potential.

SPEAKER_03

It's a huge piece of it. That's a huge piece of the puzzle. And look, we're seeing more startups than ever before because a bar to starting companies is lower. And the bar for existing companies that are already kind of incumbent or larger to add adjacent procs is lower than ever before. So net net is that there's a ton of stuff happening right now and there's lots of noise. And there's two ways I think about one would be we've always invested in deeper technical kind of modes, and you can argue kind of how deep some of these technical motes are, but we've invested in stuff like in generalist AI, which were started by a guy from Boston Dynamics and two folks out of Google Deep Mind to do a foundational model for robotics. I just did a stealth investment in the AI data center security space, which required expertise in GPUs in networking, in software and security kind of all at once. So, you know, those types of things I think are are areas that kind of will not get run over. Today we just announced a 57 million total round of funding, well, two rounds from a company called Surf AI, which we did with Iliad Cyberstarts and Philippe at Excel and insane, insane team. And what they're trying to do is provide this contextual layer to connect all these different silos of data together, not just security silos, so that agents can continuously look for risk, close the gaps, and help people remediate that risk. So, you know, these are some pretty big ideas. And I think that those kinds of things aren't gonna go away with a skill tomorrow, frankly, right? Because it's much more deeper technical workflows, a deeper understanding kind of the domain. And I think it requires special founders as well. When I think about it, I was talking to my partner the other day. If you're gonna invest kind of in an area that she thinks could be run over in six to 12 months, the question then is if you think about it in an analogy like this, imagine that AI is this insane jet stream. You meet a founder, you drop them in the jet stream. Are they gonna struggle to keep up with the jet stream or are they gonna be in the front of the jet stream, constantly shifting and adapting? Because you know, I think motes that we thought about back when I started investing 30 years ago, you're talking about 12 to 18 months to recreate software. Now it could be days, weeks, months. And then ultimately over time, is that if you're a founder that you can consistently navigate and be ahead of the curve, continue to ship new product, continue to adapt. And perhaps you have different motes. I think moats couldn't be different right now. It could be community modes, it could be, you know, networking motes, it could be building a data flywheel from every user that you get on board to kind of build that. But look at the end of the day, you got to invest in amazing people. And if you are gonna get into that space, you better make sure damn well that these founders are gonna be outrunning everyone and more importantly, know when to get out of that jet stream because you know, because I think that's that's part of the whole game too.

SPEAKER_02

Yeah. Are these moats and this kind of platform shift as what you consider kind of the autonomous enterprise? I know you speak a lot about the autonomous enterprise.

SPEAKER_03

Yeah, and and I kind of, you know, we were just thinking about kind of a marketing term to kind of launch our fund seven back in July last year. And I didn't want to use the word AI. So I just came up with this theme thing called the autonomous enterprise. And and the idea behind that is that, you know, we all believe and wish. I know Max that start a one-person company that's doing $100 million of ARR and there's just agents running around and you don't have any other people. And and just say that's a really amazing kind of aspiration and goal. But when you look at the enterprise, when you look at banks and JP Morgan and City and all those places, they're not gonna actually get to that place. However, all the primitives to make that happen, uh, the core infrastructure, all of that and the security uh thinking has to be rebuilt when you have a world where each of us has 10, 20, 100, 1000 agents working for us all the time. And you're starting to see that now, kind of seven months later. What's it like to have agents at scale doing work all the time? You've got to create a new security model. We invest in a company called Keycard.ai with Andreessen, and they're actually giving agents their own identity for their own authentication and authorization. And that stuff is given to people at runtime. It's not forever. So, for example, when we log into Salesforce, we may have those credentials forever, but agents may just need to do a job and go away. That requires a new way of building things, right? So I think that's been another core part of the thesis is investing in that whole underlying infrastructure around that. And I think there's still gonna be tons of opportunity because as much as people are vibe coding things and and doing other things like that, the last mile is really, really tricky to deliver these things at scale, privately, securely, and everything else.

SPEAKER_02

Shout out to uh Amanda Robson, who I believe is in Key Card with you as well, right? Okay, yeah, she's great. She was on the pod. Definitely check her episode out if you're listening to this. Our LP based spans from individual operators to institutional allocators, and Angelist has been instrumental in supporting all of them. They handle everything from investor onboarding and accreditation to distribution and tax documentation, creating a seamless experience across geographies and fund types. Plus, all of this is available on a single modern platform. For an LP-based IRs with over 300 C-suite and VP level operators, this kind of white glove service and seamless workflows is so important. Also instrumental that we support our institutional LPs that we're fortunate to work with, and Angelist is able to do so every step of the way. If you're looking for a platform that can support any type of LP investing in your fund, learn more at Angelist.com slash GTM fund. And yeah, back to that kind of agent point. That's a super interesting use case. I think we're kind of hitting that agent wave right now. I had another tweet of yours pulled up here. Forget employee of the month. Companies should start doing agent or skill of the week. The hardest part is not the tech, it's helping people see what agents can actually do. Show the skills, make it visible. That is how organizations change. So many more ideas on how to get your agent, uh, your org agent red pilled. What are you seeing inside of your companies? How are people leveraging agents? And is this an area where you know you're expecting as an investor companies to hire a lot less headcount? Are you expecting profitability sooner? So maybe this is like a very two-part question. But as we get into the agentification of a lot of these businesses, where do the expectations go from the investor side?

SPEAKER_03

Yeah. First, let's start with uh the first principle. If you're starting a company today, I want to ask the founder how much of your code was written by agents? And uh let's just think about a perfect world. Zoom out two years from now, you're successful. What does your company look like? Right. And hopefully the answer is is got you know red-pilled employees, they're they're all over agents. You know, we're kind of wired our system together so that we can spend more time doing the creative thinking and the creative selling and things like that. And when I look at all the startups we backed in the last year, I'd say almost all of them are shipping at least close to 100% of the code, depending on how complicated this stuff is. If it's going into more of the robotics layer, there's probably less to happen there. But I'm talking about your general software, it's mostly all the codes written by agents. And then what's happening is that people are using it for product marketing, right? You release ship, you release features, you ship features, then you start having uh agents starting to review the features and uh thinking about how to help your salespeople understand what those features are, help you think about what release schedules are, and just kind of create dashboards around that. So I think what I'm looking at is back when I first started the bottleneck, even up until December, when we all got clawed pilled, right? With Claude Code and Claude Cowork, the bottleneck was always engineering, right? I mean, Max, you're like talking to sales and marketing people, they're always like getting ready to ship the thing three months from now. It's been three months to kind of get this release out. And how do we do pricing? And how do you make sure you sell this thing without screwing up your existing installed base or making sure your salespeople aren't kind of previewing something that they won't want to buy and kill your quarter, right? All those things that's pretty much gone away. Now the bottleneck is not engineering. Engineering's almost shipping too fast. I'm looking at organizations now where the tech is just constantly spitting stuff out. I'm not saying that all of it's good stuff, right? There has to be some quality kind of filter around it, but let's just say in general, engineering and code is not the bottleneck. It's actually what people, what cut what your employees can absorb, what your sales people can absorb, what your marketing people can absorb, what they can kind of market to their customers and how much your customers can absorb. And so that's kind of why I've been talking a lot about this, because then the thinking is that if you're an incumbent, let's just say 100 people, 200 people, 300 people, if you're not doing it to yourself right now, then 30 other startups are doing it to you as well. Yeah. And eventually they're just going to outrun you. So basically that's the pain that everyone's looking at right now. I mean, every single company, every board meeting I've had in the last few weeks is not only is how much of your code is being written by agents, not AI agents, it's also what other functions are using agents and how are they using agents and how are you delivering and how are you moving with speed? And so that's kind of why I keep bringing this up. And ultimately, you know, you know what claude skills are, right? They're basically, I think every task gets codified into a markdown file where it's just a skill. A skill could be, you know, analyzing a balance sheet. It could be, you know, moving, you know, doing double entry accounting. It could be kind of when someone makes a post, you know, how do you promote that? Right. But these are all skills and that becomes codified. Your business and how you run it becomes codified in these markdown files. And then, you know, imagine having a company that has thousands of markdown files and agents kind of point at them and pick them up and then just kind of run off and do work. And I think that the startups are actually doing it first and people are catching up.

SPEAKER_02

Well, talk about Clay in a little bit, which had kind of this like almost like flatline, and then all of a sudden figured it out like six years in. And I'm sure you have companies that are in this almost other bucket where they were hockey sticked in the 20 teens, early 2020s, and now they're kind of flatlining and in this kind of no man's land. What are you seeing as like best practices or what are you telling these companies to do that are kind of like in this no man's land of maybe they got to 50 or 100 million or something more ARR, but they're not AI native. And maybe to get on parity with some of the up-and-comers in their space that are starting to take business or grow quickly, they would have to tear it all down and you know, start over with kind of the experience and knowledge and the customer base that they have, but like almost a new product. Like, how are you working with those types of companies?

SPEAKER_03

Yeah, I mean, th those are the discussions to have. And what I like to do is point to kind of what's been successful. I mean, look at Intercom. I think they're a great example. Intercom. Yeah, they they burned the boats down early.

SPEAKER_02

Airtable was another one that basically rebuilt the entire product, right?

SPEAKER_03

That was they did. Well, just the founder, right? And so I and then look at Atlassian. I think that Alassian said they hired they're gonna let go of 10% of the people, but I think more importantly, underneath that all, I think they're gonna uh release their CTO who wasn't built for the aging native world and promote two younger people. And then another perspective would be let's look at Snowflake. Frank Slootman left, he was a great sales and marketing kind of genius, had insane performance, and is like, look, uh, I'm gonna hand the company off to someone that's an AI product person because we need to have AI product vision. And they took a little pain. That company, you know, for example, bounced back. And so I think that first and foremost, you have to make sure that the leadership is not only fully behind becoming not AI native, agent native. And then two is if they're not and they're not long for this world, you have to make sure that you have people that are. And wrote about last week, you've got to not only lead from the front, I call it the sandwich model, lead from the front or the top, but in order for it to stick, it has to be organic. It has to be built from the bottom up, which is why I said, for example, companies will say, Hey, I get this agent stuff, but how do I even get started? Right. I think the hardest part is that not that people don't want to actually use agents, is that they're not creative enough to actually think about what to do. So if you have people that are peers that might be in different departments, uh every week share, hey, these are two skills I created or things that I automated to give people that light bulb, aha moment, to give them the agency to do it. I think that kind of opens up lots of possibilities. And you can see that becomes contagious over time. But it's just a leadership saying we must do this, this is how it happens, or you're fired and set mandates. I think you're gonna set up for the wrong um criteria, right? So I think it's kind of both. And then ultimately I think the companies are gonna win are gonna have more product-driven, kind of AI-driven leadership, the ones that will succeed and navigate this era. Because if you're more of a sales and marketing-oriented CEO, I think it's gonna be much harder for you to lead the charge around that. So that's kind of what I'm seeing, you know, and and so that's what you see from the largest companies in the world and also from smaller ones as well.

SPEAKER_02

Yeah, they have to adapt, and sometimes that could be a violent adoption there where there's you know a riff or or you know, bringing in different people that have different skill sets that are more AI native. But you know, that's not just on the engineering side, too. That's on the GTM side of the house. That's you know, it's it's across the board, right?

SPEAKER_03

It is, it's 1000%. I mean, look, at the end of the day, I I told you kind of what we look at when we talk to founders. Like you would imagine that all of them are gonna use agents as much as they can. And but you know, it happens. I was interviewing a VP of finance for one of our portfolio companies, and the first question I asked was, Hey, what AI are you using? How do you use it? And if they're just using kind of co-pilot on, you know, in their Microsoft Edge browser or whatever. But if they're actually spending time kind of spinning stuff up with Claude or using cowork or just experimenting with it kind of on the weekend, just like we talked about, Max, is that you have to be a student of the game. And I think the market is moving so fast, there's not enough time to actually in the day to keep up with what's happening because you have to do your job. But I think the people that are curious, you just have to be curious. You said put your hands on the keyboard and mess around with stuff, whether you're either whether you're a VP or whether you're an individual. I was talking to a head of sales the other day saying, What are you doing for AI or head of marketing? Head of marketing told me, Hey, I was gonna hire seven people, I just hired two instead. And when the first hire I made, the dude is just automating all my stuff right now, right? And so that's kind of the energy energy that you really, really need, right? So not only your product has to be agent native, but your company has to be agent native as well. And I kind of kind of and the more you become that, the more opportunity you'll have and the faster you can go and and the more you can do.

SPEAKER_02

And that starts with founder DNA, first and foremost. So that's what goes back to your you know, P and people, your testing for hey, you know, how uh AI red pilled are you, and how much of your code is agentic. And okay, so I want to pivot there then to GTM. So yeah, uh, let's talk about Clay. We're gonna give him some good air time today. How'd you find them in the first place?

SPEAKER_03

I give my partner Elliott all the credit because he's the one that's closest to it, but I've obviously spent time with both founders because we're a small shop. Uh basically Elliot met through a friend of his and box, I think, had invested in these uh two guys previously, and they're starting something new with Box. And we had sent Box uh before that customer and superhuman. They were grateful to be part of those in the initial inception round. So that's kind of how we uh how we take care of each other, you know.

SPEAKER_02

Yeah. So what was it like in the partnership with them kind of in the early years before they kind of found their footing? How do you, as a VC, you know, be a good partner?

SPEAKER_03

Yeah, and by the way, the running joke is that the first time they came in within 20 minutes, Ellie and I looked at each other like we can't let them out of the room, we have to invest. And we caught them on a Friday afternoon, like at 5 30 before they're heading to the Silicon Valley. And we knew that if we didn't lock them down, then you know, we may lose the opportunity. Was this before Varun joined? Yes. Okay, yeah. Yeah, this this was the original founders were Kareem and Nikolai. And we just had Ellie and Hyde and Vulcan Mindmilled as we just said what they're building is is super fascinating. I put some links up to the original decks kind of in one of my posts, but the running joke I kind of joke is that they're almost building kind of air table, but a more programmable air table back in the day. But the running joke is that air table won because no one wanted a more programmable air table because it was programmable enough. So that that's the running joke. But then over five or six years, you know, I think Sequoia came in for a fast follow, like 15 months into it, and led a you know, led an Around. And one thing the guys never did is they never burned a lot of cash. But it's like, great, I'm gonna put that on the side, save it for a rainy day. And they were just itering and itering and entering and entering, just making the product. It's not like they ever stopped. They're always testing new things, always trying new things. And you know, it's one of those things that I have another framework, and I think Elliott executed this to perfection. I call it the three C's or the three CHs of working with founders. One would be you have to know when to cheer. And I think knowing when to cheer is when the shit hits the fan because it always does, when they lose a big customer, when they lose around a financing or whatever, you gotta peel them back off the ground and say, hey man, you gotta keep going. Like you have this vision, you're just there, it's just one step back, keep going, right? And sometimes, you know, founders psychologically just get beaten up. It's a tough job. And the other I like to say is you have to know when to challenge. That's when founders feel invincible. We're just doing so great. Let's burn, let's triple our burn, let's keep going, you know, like and they don't see they don't have any blind spots. And your job then is to challenge them for blind spots. And the third one, I think my partner Elliot, you know, kind of leveraged to perfection was knowing when to chill. Sometimes you just let them do their thing, like trust them, let them do their thing. And he would meet with them every six weeks, just see what was happening, what they're eating. They were, but they were never stopping. And so that was, I think, the most important part is they had the breathing room from the people involved to keep cranking around to experiment, and they kept their burn really, really lean until they figured it out.

SPEAKER_02

Having three kids that are under six, I resonate with all three of those CHs. It works the parenting also to a certain degree.

SPEAKER_03

Yeah, but it's hard sometimes when you're in the grind, dude, like psychologically to actually take a deep breath and not do the opposite. Because the tendency usually is to do the opposite when those things happen.

SPEAKER_02

Yeah, that's why it's good to have a partner, whether in VC or uh, you know, in in parenting. Yeah, like okay, then I know I'm supposed to chill or cheer right now, but I really want a challenge and I'm gonna vent to you instead. And then, you know, that that's good. All right.

SPEAKER_03

So you want to know what happened, how they ended up kind of where they were. So they did this for quite a while, and then they started seeing use cases. And, you know, sales folks for prospecting started using the platform saying, Hey, this is pretty awesome to do do prospecting. Can you start adding more data feeds into your platforms? It's all in one place, and then you have this air table like format, and you make it more programmable, and they started doing that. And lo and behold, you know, Kareem's like, let's go do this. And Brune got started getting involved. Let's go do this. So they focused on it and they leveraged their core underlying platform. They end up doing 600k uh year one, and then year two, there's 4.6 million, then they did 30, and then it crossed 100 plus million this past year, and they're well on their path. I don't know what they've announced to do much, much higher growth. But that was kind of where it started. It took a while.

SPEAKER_02

They kind of destroyed a product maxim for me that I held closely when I was at outreach, which was you kind of need to build your product for like at the highest level, B players. Because if you make your product too sophisticated, you're gonna rule out everybody else. And the A players really only represent like a small percent of you know your total adjustable market. So you have to make it easy to use. And what they did very well is they went to the agencies who are all kind of the tinkerers and the rev ops people, and they are like the the a players, so in so to speak, in a lot of these organizations that spun out and started these agencies for a reason. And then those agencies would sell to customers, and so Clay would, you know, essentially would sell into those customers and then be used by those agencies, and they ended up being a massive go-to-market for them. And now it's kind of, you know, they were able to evolve from there from the you know very early days to a platform that kind of everyone can use. And I spoke to Farun, and I think that's kind of the direction they'll continue to go in. AIL obviously enable, you know, more kind of ease of use there as well. But and then they just had the ads product come out, so you're seeing them enable kind of sales-led growth, product-led growth, and ad-led growth as well. So there's they're doing a good job of kind of adapting.

SPEAKER_03

I think is an absolutely massive TAM to only spend money on the prospects that you've been tracking that sync directly to um your sales for us or your hub spot or whatever you're using, I think is is insane. But yeah, Max, you you nailed it. I mean, when they started this idea, we'll call it the cladency, and they've been writing a lot about it. It was kind of like, wow, that's that's a unique angle. It was almost the equivalent of um outsourcing kind of the idea for deployed engineers before FDEs became really cool. So you had these bespoke kind of cladencies, which could be one-person shops, two person shops who are then sort of making million, two, three million dollars a year working on implementing clay. And sure, I could implement clay, but even when we kind of did some stuff last year, I used a clagency to just speed up the process because I just didn't have the time to kind of think through it all. But I think that was definitely uh one mode. And then what they did with then was that they had a community round of funding where they allowed some of the folks who are their most loyal advocates to also invest and then can also exit as well, kind of as a company kind of grew. So they did a lot of other things from a moat perspective beyond just tech, because you know, we talked about earlier tech is kind of the thermal. And I think the second thing is they continued adding more data feeds, and now they're in the jet stream. They're continuing to add more products, right? Because they have this installed base, they added this ad. You know, that probably wasn't on their radar 18 months ago, but they're smart enough to think about kind of what else do my constituents want. How do I do it in a unique clay way? And and they're continuing to run faster and faster. And they do feel the heat. Trust me, everyone's saying once you get to a certain size, everyone's like, I can do this myself. I can, I can write. I mean, just it's just the nature of the beast. Um, but I really have a lot of confidence in Kareem and Varun and the rest of that team. It's it's been really amazing to watch them continue to innovate and ship at speed. I'll say ship quality stuff at speed. It's not just shipping at speed, it's shipping with quality quality stuff as well.

SPEAKER_02

Excellent. Well, what we could nerd out on clay quite a bit here, but I know we gotta we gotta jump off soon. So last thing that I want to talk about with you is you're doing anywhere between 500K and $30 million checks into companies right now, and you invest and look to invest for your first check at inception, right? So it's a a wide gap there. What's your kind of fund model? What's your thesis? What's your kind of mantra as you go to market with a fund like that?

SPEAKER_03

I think I'll make it it's 500 to 15 million, but which means we could probably co-lead 30 million rounds, dollar rounds, which we've done in the past, right, with Tesla, which is Guy Pajarni's third company. He's the founder of Sneak and helped, you know, get that started and well over $300 million of AR right now in the security space. And but look, at the end of the day, um, our model is we want to make sure that we have it's a $250 million fund right now. We could have raised a lot more, we just don't want so much capital. And I'm a believer that there's two kinds of ways to make money in this world. It's basically you go big and you see all the people going bigger and bigger and bigger. You go niche, specialize, or you go home. And in our case, we specialize by stage, which is inception or literally first founding kind of investor uh with partnering with founders, and kind of we'll call it the autonomous enterprise. And that's a broader play. It's not just kind of enterprise infrastructure and agents, but also that includes physical AI. So we've done stuff in bio AI, robotics AI, people doing my manure discovery, we'll look at some space stuff. So, but they're broader, broader kind of thesis, but it's a lot more infra-heavy. And uh, and that's kind of what we're doing. And so I think different founders require different checks and different companies and different ideas require different check sizes to basically recruit talent, right? Because ultimately it comes down to I think capital used to be constrained. I think talent is more of a constraint, getting really great talent. And you know, when you find founders, you want founders who can be Pied Pipers so that they can actually bring talent from all their other places. They're great leaders and people follow them. So, like in a case like Guy, when he's out starting his third company, he could have raised 50 to 75 million dollars right out of the gate from anyone. I mean, he had people kind of coming after him. And we've just had this strong relationship for a long period of time. And, you know, he wanted to go on the journey with us, and we ended up co-leading a $25 million round with GV. And then they raised another hundred six months later from indexing Excel. And there's other ones, right? I mean, Surf AI, we ended up partnering with Gilly at Cyber Starts on that one, and then uh Excel just came in and we now have 57 million of funding on that. And you know, that was a bigger check. These are third-time founders, five actually super, super experienced founders. But at the same time, we have a Joao from Career AI, he only raised two million dollars. He was a one-man guy in Brazil as he was getting his open source infrastructure product started. Then he moved to San Francisco and we ended up putting more money in kind of an interim round, and then he had an A led by Insight. So there's all different kinds of flavors. First time founders may have less, third time founders may have more. But we want to make sure that if you want to win and have ball control, which is meaning that we want to lead deals and price the deals and join the board, you have to have the flexibility to write small checks and make it feel meaningful to the founder that you're going to care deeply, and also write bigger checks because you don't want to invest 15 years into a friendship. And then when they become ultra successful, you can't actually write the check. Yeah. And that's kind of why we pick the size that we have right now. And that's, I think, our model. But look, there's lots of other models to win.

SPEAKER_02

So you've been writing your newsletter, what's hot in enterprise IT and VC for years now. If you had to summarize it today in one sentence, what is hot in enterprise IT and venture right now?

SPEAKER_03

I think that what is hot right now, besides kind of everything agents, is the fact that everyone is feeling like this is the biggest opportunity that they've ever seen and will ever see in their entire lives, but also feeling absolutely horrified and scared every second at what is happening and trying to capture that emotional aspect every week because things are moving so fast. I think is kind of capturing that emotion and that thought process, I think is kind of what is top of mind for me right now. And I know it's top of mind for a lot of founders and top of mind for a lot of investors, because that's what we talk about every day. It's like on the one hand, you have things that were once great that are getting killed, and on the other hand, you have things that weren't great and now are killing it. It's like what is going on right now? Like, how do you kind of maintain sanity? I think that's what's really hot right now.

SPEAKER_02

Love that it resonates deeply with me. You know, we're in fund three right now, and um, certainly understanding of ball control after fund two, going slightly bigger with fund three. So, yeah, I mean, all resonates. And who would have thought after 30 years of investing and doing pretty well and also sharing so much of your knowledge with everybody through Twitter and and Substack that you would have all these founders, amazing, incredible founders that would want to come to you and you know, say, hey, we'd love to take your money and you know, the next company that I do. So it seems like a fantastic model that you're executing to a T. And you know, I speak for all of us out there on the interwebs. We appreciate all the kind of sharing that you do through Substack and Twitter and LinkedIn and whatnot. So thanks so much for coming on the show. This was super insightful. Love the five Ps, the three CHs, everything else we talked about. And um, yeah, thanks a lot.

SPEAKER_03

Hey Max, thanks for having me and and uh love the content. Keep it cranking because I I do learn some things from uh GTM Now as well.

SPEAKER_01

That was another fantastic episode of the VC series on the GTM Now podcast. Head over to Apple, Spotify, or YouTube and give us a like and subscribe, and we'll see you on the next one.