How Browserbase Raised $67.5M to Power AI Agents & Vibe Coders

Paul Klein IV, founder & CEO of Browserbase, announces their $40M Series B at a $300M valuation led by Notable Capital. Browserbase has raised $67.5M in total funding in just 15 months (!!) since founding the company. Existing investors include big Silicon Valley funds like Kleiner Perkins, CRV, Basecase, and more. Paul shares how Browserbase is powering the next generation of AI agents and “vibe coders” by providing programmable, cloud-hosted browser infrastructure—what he calls “Twilio for headless browsers.” Special thank you to John at Strike Capital for the introduction, as well as Bucky Moore and Turner Novak for some incredible questions. This is an awesome interview. We cover: * How Browserbase works and real-world automation use cases * Round dynamics, fundraising strategy & investor support * Their all-star angel network: Jeff Lawson (Twilio), Patrick Collison (Stripe), Guillermo Rauch (Vercel), founders of Okta, Stitch, WorkOS, and more * The evolution of their brand (emoji story) * Go-to-market strategy and scaling past 1,000 customers * Launching Director.ai for “vibe coders” * The future of AI agents, authentication, and automation * Paul’s solo founder journey and team culture in SF Whether you’re building an agent, automating legacy workflows, or just curious how infra startups are scaling fast in AI, this one’s for you. Check out Browserbase: https://www.browserbase.com/ Molly on X: https://x.com/MollySOShea Paul on X: https://x.com/pk_iv Brought to you by: Brex—The modern finance platform, combining the world’s smartest corporate card with integrated expense management, banking, bill pay, and travel. As a Sourcery listener, you can unlock up to $500 toward Brex travel or $300 in cash back, plus exclusive perks to help you move even faster: https://brex.com/sourcery Kalshi—The largest prediction market and the only legal platform in the US where people can trade directly on the outcomes of future events: https://kalshi.com/sourcery Fourthwall—The #1 way to sell merch online — Fourthwall is the easiest way to launch a fully branded merch store—used by big brands like MKBHD, Acquired, & even the Smithsonian. 100+ products. No upfront cost. Visit Fourthwall to start today: https://fourthwall.com/ Follow Sourcery for the latest updates! https://www.sourcery.vc/ TIMESTAMPS: 00:00 - Series B Fundraise Announcement 02:21 - Introduction to Browserbase 04:04 - Use Cases & Applications 07:56 - Customer Growth & Product Development 10:48 - GTM & Branding Strategies 22:53 - AI Agent Ecosystem & Authentication 28:08 - AI Regulation & Deep Fakes 30:07 - Launching Director AI 30:54 - Empowering Vibe Coders 33:52 - Customer Service & Team Composition 41:28 - Why Founding a Startup Should Be Last Resort 44:38 - Managing Performance As A Solo Founder 48:38 - Future Outlook & Kalshi Predictions #Browserbase #AIInfrastructure #PaulKlein #Startups #Automation #SeriesB #AIagents #DeveloperTools #VibeCoders #podcast #investing #technology #venturecapital #entrepreneur #startup #siliconvalley

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  Drop this pod, you know, the day we're announcing our series B fundraise, which was led by notable capital Glenn Solomon, joining our board. This is a company that's only about 15 months old. Jeff Lawson, the CEO and founder of Twilio, Girmo from Bursell, Patrick Halsin from Stripe, I feel like I have Thanos rings of DevTools people that I really respect. Great thing about Bursell, Twilio and Stripe is they all built category defining companies. That's what I seek to build with browser base.

• 0:22Molly O'Shea?

  Forty million dollar series B. What does this bring your total funding to?

• 0:25Speaker 00

  Sixty seven and a half million, more opportunity to grow our business, help more customers and really double down on this crazy idea that AI is going to control your browser and do tasks on your behalf online.

• 0:36Molly O'Shea?

  Paul, welcome to Sorcery.

• 0:37Speaker 00

  Thanks for having me. I'm happy to be here.

• 0:39Molly O'Shea?

  I'm so excited to have you on and I love the merch. You know, I love merch.

• 0:43Speaker 00

  We got the Bs on, series B, browser base.

• 0:47Molly O'Shea?

  B, that's right. Congratulations.

• 0:49Speaker 00

  Thank you.

• 0:50Molly O'Shea?

  Could you share more about the round?

• 0:51Speaker 00

  Yeah, absolutely. I mean, we didn't just get the hats for the round. The company is browser based, but you know, we're excited to drop this pod. You know, the day we're announcing our series B fundraise, which was led by notable capital Glenn Solomon, joining our board. And you know, we're excited. This is a company that's only about 15 months old and to go from zero to, you know, four rounds of fundraising. And also I got married, you know, it's been a pretty busy year and it's just more, um, more opportunity to grow our business, help more customers and really double down on this crazy idea that AI is going to control your browser and do tasks on your behalf online. And we get to play a small part of that.

• 1:28Molly O'Shea?

  And this was a forty million dollar series B?

• 1:31Speaker 00

  Yes.

• 1:31Molly O'Shea?

  What does this bring your total funding to?

• 1:33Speaker 00

  So all in, it's sixty seven and a half million, though I kind of wish we just raised the extra 500K. So I didn't need that extra half, uh, but no, sixty seven and a half million. And we have a bunch of great investors, Kleiner Perkins, CRV, Base Case Capital, as well as a bunch of great angel investors, just everyone I really admired and looked up to is now in some way involved in browser base, which means we have a good team around us to keep doing what we want to do.

• 1:56Molly O'Shea?

  It's amazing. And before we go too deep, I do have to say thank you to John from Strike Capital for the intro and also Bucky Moore, uh, was a previous investor at Kleiner. And he gave really great questions. So we have some awesome questions.

• 2:10Speaker 00

  He's got the inside scoop, so.

• 2:11Molly O'Shea?

  Yeah, I know a lot about your personal life.

• 2:13Speaker 00

  There we go. That's good. I build in public, you know, everything.

• 2:18Molly O'Shea?

  Um, okay, so to start, can we break down browser base? For some people, the audience is like pretty diverse. So there might be beginners, there might be more technical people, there might be more finance people. What is browser based to the normal person?

• 2:31Speaker 00

  Yeah, absolutely. So when I think about browser base first, to start with, what is all of the work that we do in our day to day lives? A lot of that is in a web browser. You know, we're interacting with software, we're interacting with web apps. You might go to Crunchbase pretty often. And there's this idea that we want AI to do work on our behalf. AI is going to take some tasks that we have and take care of them for us. And if we want AI to do the same work that we're doing, we need AI to use a web browser to work the same way we do. And Browserbase is providing that web browser as an infrastructure product to developers who are building these AI features that are going to, you know, you use in your day to day life. So you think of us as a developer tool, we're an infrastructure platform, and our product is browsers at scale.

• 3:12Speaker 00

  So if you want your AI to control the web, you need to give it a browser, we provide that browser. It's kind of like a database company in some ways, right? You know, every application has these primitives, you know, maybe storing data, maybe it's like manipulating images. And also now it's controlling a browser. And not only do we provide the browser infrastructure, browsers at scale in the cloud, but we also have this framework called Stagehand, and it's the code interface for the browser, which gives developers and LLMs an easy way to programmatically control a browser. So to break it all down, if you're building some sort of cool new software, some AI agent that's going to go to a CalShe's website and place a wager, you're going to need to say,

• 3:51Speaker 00

  "Hey, AI, go to this website, find the appropriate bet, click on the bet button, and then submit my information." Those are all programmatic commands. Browserbase is the runtime for those commands. We run the browser as well as give you a lot of tools to make that easier.

• 4:04Molly O'Shea?

  So that's a great use case, but what are some other like typical like daily use cases that people use in their workflows?

• 4:10Speaker 00

  Yeah, there's a ton. I mean, you can kind of think of the primitives of work of like, you know, you're going to a website, you are filling in forms, you're probably reading data on a website, you're probably clicking some buttons, and then you may be downloading and uploading files. A use case that I used Browserbase for recently was I had to file my Delaware franchise tax. And if you've ever had I do as a founder, it is a really annoying web form on the Delaware website. Where you go in, you fill in your cap table and you correct some information and you submit that. You know, I don't think that people need to be doing all this, you know, repeated form submission all the time. They can build software using AI to do it for them and Browserbase kind of powers that. So it's a lot of the boring stuff and it's a stuff where there's probably not first party integrations. You know, if you're gonna go book a flight or find a restaurant, why not just use the OpenTable API?

• 4:54Speaker 00

  Or, you know, a flight booking API. But, but if you're going to, you know, try and renew your driver's license, the DMV is not gonna have an MCP server or some sort of API or some sort of first party AI integration. It's going to be this, you know, website that people have to interact with. And there's many, many websites that exist, both in the United States and globally, we see a lot of adoption globally, of people who built websites for people to use, but now they want AI to automate those websites. And people are using Browserbase to automate, you know, getting quotes for lumber in Mexico or finding all the 7-Elevens that are open in Japan. These are things that there's not gonna be first party integrations available for. And Browserbase is that primitive that powers all of it.

• 5:33Molly O'Shea?

  This might pop up a couple of times during the conversation, but what does MCP mean for people that don't know?

• 5:39Speaker 00

  Sure, yeah, that's a great, great question. So there's this kind of new way that people are exposing tools to AI. And the Model Context Protocol is one of those ways that you can standardize and give your AI a set of tools to call. So Browserbase has a way to expose the browser to your agent via MCP or MCP server. But more and more developers are trying to find ways to integrate more functionalities to their AI agents. And, you know, Browserbase's browsers are one of those primitive functionalities. And the MCP is just one way to integrate it into your software.

• 6:11Molly O'Shea?

  And you've also mentioned that Browserbase is like the Twilio for headless browsers. So what does that mean?

• 6:17Speaker 00

  When I think about Twilio, I think of a company that built APIs for telephony. You know, you can use Twilio to programmatically send a text message or programmatically make a voice call. And a lot of those primitives carry over to Browserbase. You know, we're building an infrastructure company that developers can use to programmatically go to a website and fill in a form or programmatically collect data off of a website. These are primitives. That people are building into applications. Twilio sold to developers, we sell to developers. There's a lot of similar business models there and I'm very lucky to have Jeff Lawson, the CEO and founder of Twilio, as an angel investor and he's been instrumental in the growth of this company and a big advisor to me.

• 6:51Molly O'Shea?

  That's amazing. Do you have any other kinds of advisors that have helped you along the way?

• 6:56Speaker 00

  Yeah, you know, I think Guillermo from Vercel has also been a very inspirational person for us and we're lucky to have them both as a customer and an investor. Patrick Hallson from Stripe, another angel investor that we're really happy to have on board. I kind of feel like I have Thanos rings of DevTools people that I really respect and, you know, the great thing about Vercel, Twilio and Stripe is they all build category defining companies. And that's what I seek to build with BrowserBase. I want to build a company that defines a category, in our case browser infrastructure, and really builds everything within that category so that developers can use us to build these primitives into their applications. So, you know, as a kid, when I started programming, I was building little software applications and I wanted them to do stuff for me. And that was really hard. I wanted to like look at my practice schedule on the school website.

• 7:39Speaker 00

  And developers want to interact with software in the rest of the world and the way to do that up until now has been very painful, but with BrowserBase and headless browsers, we give developers the ability to control a browser and really interact with the rest of the internet. And that unlocks a lot of really cool software opportunities.

• 7:56Molly O'Shea?

  I heard that you have over 700 customers. Is that right?

• 8:00Speaker 00

  We actually just crossed 1,000.

• 8:02Molly O'Shea?

  Oh, my God. You've crossed over 1,000 customers.

• 8:05Speaker 00

  It's really exciting. It's a lot of mouths to feed, but for our 25 person team. But we're really thankful for it and our product led growth approach makes it so that, you know, every customer can self service. They can sign up when they grow. They can talk to our sales team. But, you know, it's been really cool to see customers start with us when they're like maybe a pre seed company. And then now we had a customer of ours just raised their series B as well. So that we can be with them along the whole duration of the journey. And our scalable model does make it quite nice where we can service a lot of customers with a lot of low friction.

• 8:33Molly O'Shea?

  What's the typical profile for your customers? Does it span a range? Like if it's self serve, is it a lot of consumers? How does it work?

• 8:42Speaker 00

  Yeah, they're all businesses, all startups. So we're very much B to B, mostly developers. Customer profile browser base has really changed over time. In the early days, when you know, people were just kind of building on the edge of AI in early 2024. It was a lot of startups, you know, it was a lot of people and hey, we were a startup ourselves. So we didn't really earn the right to go get a larger customer until we grew more and really proved out our infrastructure. These days, we see customers from like all different types of business sizes, from you know, still the early stage startup that's innovating to maybe the scale up that has a thousand employees and really wants to double down on some internal automation they're building. Or the publicly traded company that already used some sort of a, you know, headless browser technology in their cloud and wants to look to what's new and what's innovative.

• 9:27Speaker 00

  So it really has spanned a lot more different types of companies as we've grown. But I think that's very traditional for infrastructure companies. You prove yourself out to these early startups. And then over time, as your platform becomes better, you win the right to work with larger companies that can trust you with their cloud infrastructure.

• 9:45Molly O'Shea?

  Sorcery is brought to you by Brex, the financial stack trusted by more than 30,000 companies, including one in three venture-backed startups in the US. Nearly 40% of startups fail because they run out of cash. Brex is literally built to help founders avoid that. Unlike traditional banks that let your money sit idle, chipping away at it with fees, Brex is designed to help you spend smarter and move faster. Their all-in-one solution combines checking, treasury, and FDIC protection into one powerful account. You can send and receive money globally at lightning speeds, get 20 times the standard FDIC coverage through their partner banks, and even high yield from day one. With same-day and even same-hour liquidity, access your funds anytime.

• 10:30Molly O'Shea?

  Companies like Scale AI, DoorDash, ServiceTitan, Hims, Anthropic, Flexport, Robinhood, and Plaid. Trust and use Brex. Start today at brex.com/sorcery. That's B R E X dot com slash sorcery.

• 10:48Molly O'Shea?

  BrowserBase acquires customers in a very non-traditional way for an infrastructure company. Can you unpack how you determine how much time you want to spend on go-to-market versus product?

• 10:58Speaker 00

  Yeah, for us, I've really thought about what do those category-defining companies do well. And for BrowserBase, you know, we are building in a new category. We earned our spot on the market map, right, and created that. So to do category creation, you have to do a lot of top-level awareness. You have to show people this is a thing, and if you want to, you know, do some sort of automation of a browser, you should think of BrowserBase. So I think it's a lot more top-of-funnel awareness than other infrastructure companies do. If you're building another database company, well, databases are kind of a well-known problem, right? And you can go maybe more top-down, sales-led, go into the customers that are in your ICP that need that problem. But for us, we're inspiring a market of developers

• 11:42Speaker 00

  to go build. Our biggest limitation of market is actually how many people can we inspire. So on go-to-market, we spend a bunch of time doing a lot more awareness, a lot more brand development, a lot more category creation to inspire developers to get started building on us. A lot of our customers are greenfield, so they're just starting to build these features. And at the same time, I'm a solo founder, so you ask, how do I break down go-to-market and product? Well, I do a lot of both, and I'm lucky to have great, you know, teams on our EPD, our engineering product design, and our go-to-market teams, and great leaders on both of those teams, actually. But in the end, I think there's not much difference, because if you're building great product, you can put that on the market and inspire

• 12:24Speaker 00

  people. So I think they actually are very closely interlinked. We don't build one-off features for customers. We have, you know, 1,000 customers now, and the way that we do product development is like every time we ship a new feature, it's like, well, how many of our customers will this impact? And what's the greatest common denominator of like requests that we can ship in a feature? And when you have a large enough, you know, base of customers, you get good direction on where you need to take the product to really improve it. And we just kind of listen to the customers, build that, and then tell our customers about the features we're building, and they end up talking about it out in the world, too. And we get to amplify that and support them and be our biggest Their biggest cheerleader. I'll end with this, like, our pricing model is really aligned with the customer. Because we're usage based. We only win when they win. We only win when our customers succeed and they ship to prod.

• 13:08Speaker 00

  So a lot of the work I do is, how can I help my customers win? How can I tell them, hey, like, here's how you should launch this product? Or here's, here's some feedback I have after I use it. Because I really want our customers to have breakout success, so we can be powering them and hopefully play a small part in their journey to building something very great.

• 13:23Molly O'Shea?

  What are some of your most proud go-to-market strategy moments?

• 13:27Speaker 00

  Well, I think every fundraise, we do a good job. You know, I think we really are figuring out how to anchor around a single day, a single moment of trying to dominate the timeline. And that's just because it's, it's hard to get, you know, corral a whole cap table to constantly be engaged. They're busy. They have a lot of investments. We work with like, pretty busy people, right? So I, I, I try and focus all of our energy on to these big launch days. Up until recently, they've been tied with fundraising announcements, though that's probably not the most sustainable thing ever. We want to launch stuff. I really like how Brian Chesky and Airbnb does their like, launch days, and I think they really try and channel the energy.

• 14:03Molly O'Shea?

  How do they do them?

• 14:04Speaker 00

  They do like, uh, it's like, every season is like a new launch release, where they released like, this new design for Airbnb recently, and they have a whole event around it. So I think traditionally, people have done a lot, a lot more like a launch week, where you see a bunch of features come out over a week. I don't have the attention span to wait a week to see people launch stuff. I like to try and do it all in one day. So if you look at our Series A announcement, which the, the copy for that was, the next billion-dollar companies are going to be built on browser base. And that's, you know, that's now true. We have unicorns using browser base, right? And then at the, the seed, it was like, you know, I forget what it was, but it was like this very clip-driven approach. It's like, hey, like, why is AI not delivering? It's because of the infrastructure. And we have like a Roomba running into a wall. And it's like, that's, this is AI? It's like, come on.

• 14:49Speaker 00

  And then for the pre-seed, it was, um, you know, really kind of our, our planting our flag in that the future is, you know, AI doing work on your behalf. So we've always tried to have really good messaging alongside each fundraise. And I do think what we do extremely well at browser base is messaging. And I think that's the most important thing for a startup. Is is your message is the thing you're building getting across to customers? And we spend a lot of time internally drafting the copy and the messaging we use. So people understand why this thing is so important to us.

• 15:18Molly O'Shea?

  And what's the typical language for that? Like, I know some people like to go after more stunt marketing. Some people like to go after more technical stuff. In terms of videos, maybe it's lots of demo videos. How do you think about all those different kinds of media content?

• 15:33Speaker 00

  Yeah, you know, it's a lot of video. I think people love video these days, but it's also supplemented with really concise copy. So for us, messaging is so important. And we think about who are we trying to message to? It's the builders. There's everyone is building something now, right? And when they're building something, they need the right tools. And if we can offer them some of the right tools, I think that will really help them. You'll feel like they have that support from early days to build great products. Regardless, like for us, I really don't think there's like one single like silver bullet. But it's really like, how do you have complimentary pieces of content? Where it's like, maybe it's legacy media, maybe it's traditional PR. Maybe it's something that goes viral because it's a little stunty, right? We have a lot in store for the series B announcement.

• 16:15Molly O'Shea?

  Ooh.

• 16:16Speaker 00

  For the series A, we flew a plane over San Francisco and had a little banner. Yeah, that was really fun. Would I do it again? Probably not, but you know, like, I think we're always trying to try new things and see what has the best ROI. And really what like gets our customers and prospects excited about browser base? And see that we are really trying to grow this thing to be a much bigger business.

• 16:36Molly O'Shea?

  Who needs billboards when you can have planes flying over us out?

• 16:39Speaker 00

  They're way cheaper too.

• 16:41Molly O'Shea?

  Are they really?

• 16:41Speaker 00

  Yeah, they're not that bad.

• 16:42Molly O'Shea?

  Is there a whole business for this?

• 16:43Speaker 00

  There is. They make their they make their margin on the banner, not on the plane.

• 16:47Molly O'Shea?

  Okay, can you take a ride if you want?

• 16:50Speaker 00

  Yeah, I asked that. I really wanted to put someone in the plane. They said no, but that would have been pretty epic. Maybe I have to pay a little bit. The next tier of plane is where you can have an engineer sitting alongside there taking content.

• 16:59Molly O'Shea?

  Get an engineer on the plane.

• 17:01Speaker 00

  I'll work on it.

• 17:03Molly O'Shea?

  Or maybe let's scratch that. Let's get you on a dragon capsule.

• 17:07Speaker 00

  Wow, yeah, that might be a little bit, maybe the series F. We can go fly up there. That'd be pretty fun.

• 17:13Molly O'Shea?

  Is Twitter your biggest distribution channel? How do you think about the different?

• 17:17Speaker 00

  Yeah, it is. Twitter, I definitely use Twitter the most for getting stuff out there. I think it's just because I've built the audience there and like I try and direct people there. I'm trying to get better at LinkedIn. Marty Kraus, shout out Marty, just dominates LinkedIn from Pylon. He's so good at it. And I um, I think that like. I'm I kind of have always tried and been like one channel and actually the journey of browser space in 2024, we started, my goal was that when you thought about headless browser infrastructure, you thought about browser base and specifically those people in San Francisco in the AI space thought about browser base. So I really tried to go where they were, which was on X and it was, you know, at events and really trying to get into that market. Now that we're going more broad and we're trying to target more types of companies,

• 17:59Speaker 00

  we are trying to shift across more channels. We're doing more YouTube, more LinkedIn, and still trying to like find the breakout success there that we found on X. But we're definitely seeing good results and I think we just have to get better at multi-channel. And as we bring more people onto our team, they bring their own expertise in certain channels. And I'm excited about some of the success we've had there.

• 18:18Molly O'Shea?

  Can't wait to see you on Instagram.

• 18:19Speaker 00

  Yeah, that's the final frontier. Watch out, maybe traded VC will have a post on, you know, the series B day or something, right? Or TikTok. I think we have our first TikTok coming out today. So it's it's fun. I don't know if developers are using TikTok to discover content yet. I know they're using YouTube. So

• 18:35Molly O'Shea?

  They love YouTube.

• 18:36Speaker 00

  Yeah, developers love YouTube. So we've done a lot of YouTube partnerships. We've sponsored great creators. Theo, Theo Brown has done some really good stuff with us as a sponsor. So I love supporting creators. I think it's actually such a great use of, you know, marketing spend because then they become advocates. They end up giving you great feedback and hopefully you're building champions with the creators you work with.

• 18:54Molly O'Shea?

  And the word of mouth is like incredible. All you have to do is go to an event and like you're at the table and you're talking with people or whatever, but the influence is like really, really strong of word of mouth in person and all of that organic kind of growth.

• 19:06Speaker 00

  Yeah, word of mouth is probably our biggest channel of growth for browser base. When we look at the signup attribution, a one that's been growing quite a bit is actually, I heard about browser space from AI. So I'm very excited. Yeah, I think 15% of our signups came from, well, AI told me about browser base.

• 19:20Molly O'Shea?

  What do you mean AI?

• 19:21Speaker 00

  Yeah, like chat GPT or, you know, maybe they're like, hey, what tool should I use for this problem? And browser base is now in the models context and being recommended, which is super exciting because it's something that like I thought would take much longer to have like some sort of impact on our signups. But no, it's happening today and I'm, I'm still trying to figure out the right way to encourage that more. I think it's the same, you know, content playbook you use for SEO. Put a lot of written content out there, make sure it's scrapable, make sure LLMs can get that into its its context as well as, you know, making sure we have our content in other places with good backlinks. I think a lot of the SEO playbook is also true for like AI SEO. But it's it's having results on our signups and it's it's pretty cool to see. I never thought AI would tell people to go buy Browserbase.

• 20:02Molly O'Shea?

  That's amazing. That's so cool. I'm gonna have to do some searches, some prompts later and see what I can get.

• 20:10Molly O'Shea?

  In terms of the branding, we have to go back to this.

• 20:12Speaker 00

  Yes.

• 20:13Molly O'Shea?

  Your logo is an emoji. That's pretty iconic. How did you come up with this idea?

• 20:18Speaker 00

  Well, it's no longer an emoji for legal reasons. Apple does own the copyright to the emojis. But I think like owning your emojis is something that I think the the guys from Morning Brew came up with this where it's like you need to have an emoji that's recognizable with your brand. That way your logo is installed on everyone's keyboard, you know, like if someone talks about Browserbase, our logo is there or something that looks like our logo is there. Since then, the branding has grown up quite a bit and we've taken the idea of, you know, what is be the building block, a square? And it's grown into, okay, well, we stack the blocks, right? And there's Browserbase, our browser infrastructure. There's Stagehand, another emoji. I think it's the this emoji. And it's in a yellow block and they stack on top of each other. The framework works well with our infrastructure.

• 21:03Speaker 00

  So we've really kind of gotten to this idea of like, how are we building the building blocks for developers and Browserbase and Stagehand? And our newest product we're launching today, director.ai, those are all building blocks that kind of stack vertically on top of each other to help people solve this problem of web automation. And it's gotten very cubey for sure. Lot of cubes on our website everywhere. I don't have a Rubik's Cube in the office yet, but I probably did get a browser base one.

• 21:25Molly O'Shea?

  I did see that. I think it was on one of your product pages. A nice Rubik's Cube.

• 21:30Speaker 00

  There's there's many cubes everywhere, so we're going all on the boxes.

• 21:34Molly O'Shea?

  You did bring up one thing, and that's the problem with the internet. So what is the problem with the internet today?

• 21:39Speaker 00

  Yeah, you know, the internet is vast, right? There are so many websites that have been built that we still depend on that are not changing. You know, and for people to interact with these legacy websites, it's humans are going there and interacting with them. There's no APIs or connectors. And I think that there's so much of this existing internet, there's so much of this existing infrastructure that people have to depend on and use every single day, that we're probably not going to rewrite all of it. We're not going to rebuild a new internet. We're, we need to have the future of software, which is AI powered, be able to interact with the old internet. It's kind of like this analogy of, um, we have Waymos, right? It's AI driving a car in San Francisco and it's doing a pretty good job.

• 22:19Speaker 00

  But shouldn't we just rebuild all the roads for Waymo so they can drive away faster and have their own like lane and stuff? That'd be ideal, but probably unrealistic. So if you can have AI navigate the infrastructure we already have built, you're able to kind of really extend the impact of AI early on without needing to rebuild the rest of the world. And I think the thing is same, the same thing is true of the internet where we have this internet, this legacy, like GeoCities and WordPress websites and old forums are still there. And we really want to help AI bridge the gap to the legacy internet. And that's why Browspace exists is really to help bridge that gap.

• 22:53Molly O'Shea?

  How do you feel about the AI agent ecosystem?

• 22:56Speaker 00

  I think it's early. I think there's still a lot of, you know, market patterns are developing, but there's certainly breakout successes. I'm a big believer in vertical AI companies. I think they're doing quite a good job. I think there's constantly a debate around like, well, won't, you know, ChatGPT Enterprise just eat up all of this market? I don't think that's the case at all. I think that to build a great vertical AI company like a Decagon, a Sierra or Harvey, you have to not only have a great product and proprietary information about how this workflow impacts a certain profession, but you also have to have like really great go-to-market skills to go get into the businesses that need this the most. And I think that vertical AI companies are going to continue to grow. And if anything, I think there might be more specific verticals and it's been more competitive. You're going to have to compete with a company that's only doing compliance for farms, you know?

• 23:38Speaker 00

  It's not like the compliance company that's a vertical AI company might be getting, you know, picked apart and, and debundled by more specifics products. And, um, I think that's going to be really interesting to see how things. Is generalized in the vertical AI space. But there's certainly a lot more to be proved out and it feels like the past two years has really flown by. And there's a lot of success, a lot of revenue. But as things settle over the next several years, I'm curious to see how the markets shift and how different categories of vertical AI agent companies play out.

• 24:09Molly O'Shea?

  Would you classify BrowserBase as an AI agent?

• 24:13Speaker 00

  We're more of an infrastructure company, so we sit below them. Our customers are often building AI agents.

• 24:17Molly O'Shea?

  Do you think people think that you are?

• 24:20Speaker 00

  I think I think BrowserBase is a category creator, which is still trying to define and share with the world what we're doing. So no, if you're watching this, we are not, we're not even really an AI company. We're an infrastructure company. We, we help cut companies use model, but people bring their own models on BrowserBase. Their models control our browsers, right? So we're more of a primitive or a building block. And when you think about the L L M O S, the L L M is the brain. The browser is just one of the tools it needs. Same with file systems, same with like video and voice inputs, same with embedding and vector databases. These are all the primitives that help power these like agentic applications and we seek to build the default browser for AI.

• 24:57Molly O'Shea?

  BrowserBase is not an AI agent, it is infrastructure.

• 25:01Speaker 00

  Yes, we, we are the only AI company in AI Grant that spent like $0 on models while we are in AI Grant. So we're very much, you know, server, security, compliance at scale, like making sure we give you all the tools and building blocks. That's where we spend a lot of our time.

• 25:17Molly O'Shea?

  Well, let's talk about security and compliance a little bit. So how do you think authentication lays into the whole theme of this?

• 25:24Speaker 00

  It's a great question. That must be a Bucky question.

• 25:27Molly O'Shea?

  Maybe.

• 25:28Speaker 00

  How do I, how do I think that authentication lands in the future of AI? Well, I certainly think that your agent authenticating on your behalf is an important part of it browsing the web. If you want to have it go book that flight, it probably needs to sign into your United account or maybe your JetBlue person, your JetBlue account. And buy it with miles, right? But like sharing your password with an agent is not the easiest way to do things. And JetBlue doesn't have some sort of advanced agent authentication system set up yet. So one, it's like, who is positioned to build AI agent authentication? And secondly, what about all the other companies that don't have that sophistication? I think that there's some really great opportunities for authentication providers. You know, one of our investors, Okta, the founders of Clerk, the founders of Stitch,

• 26:13Speaker 00

  the founders of WorkOS, all investors in browser base. And it's because I believe that if you currently are using WorkOS, Stitch, Okta, or Clerk, they can enable agent-based authentication really easily. It's a switch because they already have the distribution. They're already doing login with those companies and sign up with those companies. Why not let your agent login on your behalf? So they're very well positioned to solve that. But secondly, for the rest of the internet, I'm really excited about like, how can we let our agent login on our behalf? And I see some really interesting opportunities with like one password. Or what if you could share your, you know, one password, password with your agent and it has some sort of limited access, limited audit access to your password vault and it can go login on your behalf and then like remove that access? So it feels like the tools are there, but the biggest thing holding back AI agents

• 26:57Speaker 00

  from really browsing the web reliably is authentication. And there needs to be some innovation there, probably from a couple of very well-established companies.

• 27:05Molly O'Shea?

  You should partner with World, Tools for Humanity.

• 27:07Speaker 00

  Yes, I think we should. I mean, I think there's like this idea of like, how do you authenticate people? And like, CAPTCHAs are really like proof of personhood.

• 27:15Molly O'Shea?

  I can't even do those, honestly. Like, I don't know where the tire ends and begins. It's too confusing.

• 27:21Speaker 00

  They've gotten very sophisticated, right? And it's challenging, but you know, CAPTCHAs exist because there was only bad bots on the internet in the past. You know, they're mostly trying to take over your account or steal your credit card info. But now there's good bots. And how do you identify between good bots and bad bots when a CAPTCHA just tries to block all bots? I think authentication helps solve that problem where, you know, your agent can say, hey, I'm working on behalf of Molly. This is her account. I'm going to help her book this flight. And at least they know it's Molly and that proof of personhood or the delegated proof of personhood will help with the abuse and the CAPTCHA problem. So that's one reason that authentication is really needed.

• 27:57Molly O'Shea?

  This goes into the topic of regulation and CalShe has a really interesting market out on this right now. So do you think AI regulation will become Federal law this year?

• 28:07Speaker 00

  Yeah, this is it's a good question. I think there will be some sort of regulation, right? Like especially around like deep fakes or video.

• 28:13Molly O'Shea?

  Yeah.

• 28:14Speaker 00

  It feels like you, if you use AI generated video to, you know, cause pain or harm or reputational harm to someone. Is that standalone AI regulation or is that just part of defamation defamation is like a law or classification? I don't think they're going to ban like models of a certain parameter size. So I I am biased on the broad yes here. There will be some AI regulation this year. But I'd have to look at the scope of the poll and see if it's all about training or something like that.

• 28:41Molly O'Shea?

  Well, I'll tell you right now.

• 28:42Speaker 00

  Please.

• 28:43Molly O'Shea?

  Um, there's about a thirty five percent chance right now on Kalshi.

• 28:48Speaker 00

  Oh, that's a good one. That's a good take.

• 28:49Molly O'Shea?

  Yeah, yeah, that it becomes federal law. But I I know your point on the deep fakes and the video like I the Google V O.

• 28:57Speaker 00

  Yeah.

• 28:58Molly O'Shea?

  V O, however you want to say it, three is insane. It's so good. And I'm watching, you know, videos on my phone on like Instagram and I'm laughing at them because it's funny. And the one of the jokes is like your grandparents are going to be like in for a rude awakening because there's these like clips of of like news reporters with like aliens in the background or like crazy events happening. And like, that's like, kind of like a ha, like silly use case. But that just mirrors like what potentially could actually be shared on the internet.

• 29:28Speaker 00

  Yeah, I think the opportunity for disinformation is really large and it's going to take some sort of, you know, proof of reality in media. The same thing along with like proof of personal, I think there's just more. I'm not a crypto person, but it seems like there's a crypto solution here where it's like there's more proof of authentication of something is legit. That needs to be verifiable by everyone. And I think there's a lot of innovation left. Like we're still so, so early in this new AI world with models and agents and all these like buzzwords and MCP. That there's going to be stuff to be figured out, but I'm optimistic that there's so many people starting companies that people will figure this out. And it's going to become net positive over the longer run.

• 30:07Molly O'Shea?

  Are there any product bets that you're making in LLMs or autonomous workflows?

• 30:12Speaker 00

  Yeah, so we're actually launching a new product today, part of Series B.

• 30:14Molly O'Shea?

  Oh, my God.

• 30:15Speaker 00

  First, first heard on Sorcery.

• 30:17Molly O'Shea?

  Breaking news.

• 30:18Speaker 00

  Breaking news. Um, you know, and it's called Director, director.ai. And it kind of like goes with this theme around the rise of the vibe coder. And what's was really interesting is I kept talking to dentists. Dentists would book calls with us on browser, saying, hey, like, we are trying to automate this back office thing. I have cursor, I have Claude code, and I want to try and do this thing. I want to use browser base. I'm like, wow, that's awesome. Like, these dentists are, hey, they're smart. They're trying to automate some software. And instead of buying software to do it for them, they wanted to vibe code their own. Because they had some, like, bespoke need, or they wanted it done a certain way, you know? And I really thought a lot about how can I arm the vibe coders to help automate their work online? And Director is that. A Director is an agent builder.

• 31:02Speaker 00

  So using natural language, you don't have to know how to code. You can say, hey, I want you to go to the Californian nursing website. Look up Julia Hu, nursing number nine, four, one, five, six, and tell me that she has her nursing license. It's my wife. And, you know, it will build you that workflow to actually go out to the website, put in the information in the form, verify this credential check, it's a common thing in medicine, to check the credential, and then give you a script that you can then integrate into your application. And what's quite cool is that it's, it helps people build these POCs where it's like, well, can AI automate that for me? Well, you can try it with Director and chat to build this automation script. And then when it works well, you can hand it off to an engineer and say, hey, can you make this ready for production? Or can you integrate this into our application?

• 31:44Speaker 00

  So we view this as kind of like, it's like the V zero for browser automation, or the lovable, or the bolt dot new for browser automation. And we're really trying to arm the vibe coders with better tools. Because we think that this is a better experience than just dropping our docs into Claude code or something like that.

• 31:58Molly O'Shea?

  Who knew dentists were vibe coders?

• 32:00Speaker 00

  Dentists are vibe coders. Uh, there's so many vibe coders. I mean, one of our customers is a 55-year-old dairy hauling company. They never hired an engineer in their entire life.

• 32:10Molly O'Shea?

  How did they find you?

• 32:12Speaker 00

  Well, they hired their first engineer this year. And that engineer used browser base to automate the aggregation of. Gas prices along a route. Previously, they had an ops person saying, okay, you're doing this route. Here's where you stop for gas. They built like a, you know, AI assistant that would go give in the route, find the gas prices and tell you, you know, what, where you should go for gas. And that was all vibe coded by this engineer using browser base. So I think it kind of has shown that AI has escaped Silicon Valley. We are, people are building agents in the middle of the United States. They're building it internationally. So it's not just our SF bubble anymore. And we think that the primitives here are really well advantaged to help power vibe coders. Cause when you're a vibe coder, you're just pulling together a bunch of primitives and trying to build your bespoke application. So

• 32:52Speaker 00

  we want to really double down on those people and the director, director.ai product is going to be the thing we do to give them better tools.

• 32:58Molly O'Shea?

  It's hit the farms.

• 32:59Speaker 00

  It's hit the farms. And the dentists, exactly.

• 33:01Molly O'Shea?

  Wow, you've hit escape velocity. I don't know where you go after that.

• 33:05Speaker 00

  Yeah.

• 33:06Molly O'Shea?

  Truly.

• 33:07Speaker 00

  To the moon.

• 33:08Molly O'Shea?

  To the moon. Love premium merch just as much as we do. That's why Sorcery uses Fourth Wall. Everyone from creators like Marcus Brownlee to podcasts like Acquired to orgs like the Smithsonian Institute are using Fourth Wall. They let you create and sell premium products without having to stress the details. They handle everything from production, shipping, customer support, taxes, even giveaways. When it's time to level up, make gear that people are actually proud to wear, that's when it's time to use Fourth Wall. And that's why we've trusted Fourth Wall for all of our brand wear at Sorcery since day one. Use my link in the description to get free credits for your first order. And for any VCs, DM me on X and I can get all your portfolio companies set up with a free samples credit deal.

• 33:52Molly O'Shea?

  Okay. So you mentioned customer service is part of your differentiator. So what does that entail for you?

• 33:58Speaker 00

  Yeah. For, for browser base, we've really invested a lot into our customer engineering team. And they're not traditional like CS people that you might find on the street. Like they're actually the most educated people at browser base. Like one went to Columbia, they went to Harvard, other went to Harvard. So like pretty smart people on our team and they're all engineers. So like, we think that when you are developer and you're trying to get support for an application you're building, you want to talk to another engineer. And that's why we've built this customer engineering team that you could hop on a call with them. You could book office hours with them. You're trying to build your application. Our KPIs, how soon can you get to production? How soon can you ship this thing? And we spend a lot of time on making that easy. As well as like, you know, we try and make the product very self-service. So for our customers, they don't have to talk to us. But if they want to, they want advice, we're here to help them.

• 34:43Speaker 00

  And it's a team that I'm very proud of because I do a lot of the heavy lifting to make, you know, when our product isn't doing a great job, which there are many parts of our product I want to keep making better. The customer engineering team really helps smooth over the rough edges because they can kind of be there with the workaround or be there with the best practice or help you avoid the landmines before you step on them.

• 35:02Molly O'Shea?

  So what is your team composition? I've heard you have a great culture, a really great team. You guys are based in San Francisco, but what's the current composition?

• 35:10Speaker 00

  Yeah, we're about, you know, I think two thirds engineering, one third go to market. So, you know, we're an infrastructure company, so majority engineers. Within those engineering teams, we have like five sub teams. And then on the go to market side, we have just one big blob of go to market people. And they all like have very clear responsibilities, like, you know, we have a sales function. We have a growth function. We have a customer support function, a marketing function within growth. But we've tried to keep things very flat. And I think that really we've tried to invest a lot in this in-person culture in San Francisco. We have an office in San Francisco. The team is in person five days a week. But it's like, we balance that intensity where it's like, hey, we people kind of show up between like nine or 10 a.m. They leave around like six or seven P.M.

• 35:53Speaker 00

  And, you know, if you step out to go to the dentist, it's like no big deal. It's not like we're building a like prison or something, right? Like, we're just trying to go back to what we had before COVID, which was like high trust, high integrity, default is coming into the office. And we built this culture that I call, we call this an emotionally vulnerable company. So we really anchor on, hey, how can we talk about the hard things? How can we talk about the feelings we have about the work we're doing? Because if you don't express that, you end up building a lot of resentment. And like a few of our values is like assume best intent, dare to trust. Like you just met this person, are you going to trust them like with their opinions? Well, you should like, that's the first thing, assume best intent, dare to trust. Embrace our mistakes. We did something at All Hands recently where we all went around and said like, hey, what's one thing we messed up on recently?

• 36:38Speaker 00

  To try and normalize that. And what engineer was like, hey, like I shipped this bug that broke production and it made me really question like, am I a good engineer? And immediately we were all able to say, no, you're a fantastic engineer. What are you thinking? But like, just being able to find those like little insecurities and validate them and then also like push people to continue to take risks and we are going to ship bugs. That's okay. It's part of the journey. But the more that we can be aware of how we're feeling about the work we're doing, the better we can work together as a team in person. So people really like our culture. We have a great video about our culture on our website on the careers page that really kind of like, to me, speaks to what it feels like to work at browser base. And the most important value is pressure is a privilege. And I got that from the US Open. We're like, you walk on the US Open, it's out there. And I love that because we were a team of former founders. I'm a former founder

• 37:23Speaker 00

  of a company that didn't find product market fit. And it's important to remember, like, when you feel customer pressure, when you feel pressure from investors or from the world, that means you're doing something right. That means that you have the opportunity to build something and people are depending on it. And I really like to try and reframe that, like, when we feel stressed out and pressured, like, oh, my gosh, that means something is working and we shouldn't be taking that for granted.

• 37:42Molly O'Shea?

  Bucky did mention in our notes that the ownership mindset is like a critical value of the company. But how did you actually come to that? And why do you think that's so critical?

• 37:52Speaker 00

  Yeah, well, I think that's the only way you can scale well. Like, you can't, you could hire a bunch of like midline managers to micromanage each team and like make sure things are on the right path. Or you can empower people and say like, hey, if you build something, you need to see it through and make sure people use it. Like, you don't just throw it over the fence to the growth team and like, hope it gets used. Like, I always try and get the engineering team to question like, okay, I shipped this thing. Are people using it? If not, should I be tweeting about it? Should I be helping grow it? Should I be going to the Slack channel telling our customers about that? There shouldn't be these artificial divides between like building and growing. I actually think that the growth team is here to help champion our engineers when they build things, but we also have no PMs and that's by design. Like, the engineers drive

• 38:34Speaker 00

  the product requirements from the customer needs. And I think that gets a lot easier when you have more customers because there's clearly a lot of needs. So you can just pick something off there, spend time with the customer, build the thing that works for them, and then own the outcome of like, well, how do I get more people to use this? Do I write a blog post? Do I make a video? And using our growth resources to make that as seamless as possible.

• 38:53Molly O'Shea?

  How do you filter for that in recruiting, though? And you're hiring people, like, how much of this is training versus just natural? Expertise.

• 39:01Speaker 00

  Yeah, I wish I had like a tweet that I could say that would like capture that. I think it's a lot of vibes. I think it's like you can pattern match really well. Like I've hired 25 people in person in San Francisco. We've had people move across the country to come work at BrowserBase. And it kind of comes down to a few things. It comes down to like, are they a missionary? Like I really, we're still in the missionary phase at BrowserBase where like every person who works here deeply cares about the mission. They all have built something before joining the company with BrowserBase and they say like, this is why it makes sense to me. So they're bought in on this thing and they want to be contributing to it. I also think that a lot of former founders really enjoy working at BrowserBase. We have quite a few X Y C C T Os or you know, former CEOs of like, you know, smaller companies that have come to BrowserBase.

• 39:45Speaker 00

  And those people tend to get it. They're also just great people to work with because they know the pressure I'm under as a solo founder. And they're a great support network for me within the company. So throughout it all, the way we find great people is really like having great people come to us. And making sure with clear messaging, like I said earlier, the right people tend to self-select into BrowserBase. And the more we can share our culture externally, we end up attracting the right people. And through, you know, a series of many conversations, we're able to figure out their intentions and their attitude are aligned to what we need to build a great company.

• 40:16Molly O'Shea?

  How are you thinking about growing the team with this new funding?

• 40:19Speaker 00

  We're going to double the team. So we're getting to 50 this year. We've had, uh, I haven't had a desk at BrowserBase since I started the company.

• 40:26Molly O'Shea?

  What?

• 40:26Speaker 00

  Yeah, I think I have one for like a month.

• 40:28Molly O'Shea?

  Just standing around, walking around?

• 40:29Speaker 00

  Yeah, I just feel like it calls and like I just like squat at people's desks. But we've moved like four or five offices now, I think. So we're moving to again, twice this year to expand. But throughout all that, like it's just, we want to grow the team. We want to grow engineering. We want to grow our sales motion a little bit more. I think a lot of infrastructure companies and DevTool companies are allergic to sales. It's like engineers don't like to sell. Both my parents were salespeople. So I'm what happens when you get like a gamer raised by salespeople who loves the code, right? So I think for us, we need to be best in class at sales. And sales is a really important tool for helping people understand how do they fit this into their product? And how do we think about helping you shape your roadmap over the next six years or sorry, six months?

• 41:13Speaker 00

  So really doubling down across all of the business and I think 50 is going to be a really awesome size. I think we've historically under hired browser base, so I'm excited to kind of bring more people into the fold. We actually hired five people this last month.

• 41:25Molly O'Shea?

  Congrats.

• 41:25Speaker 00

  Thank you.

• 41:26Molly O'Shea?

  Wow. That's great. That's big. And we didn't even talk about this yet, but let's get into your background.

• 41:31Speaker 00

  Oh, please.

• 41:32Molly O'Shea?

  So you've mentioned a couple of times, you're a solo founder. What? Okay. Yes, you mentioned that, but also I just remembered, I was listening to a pod with Turner Novak and he said that you said founding a startup should be last resort. So wow. Can you like bring that into the whole picture here?

• 41:51Speaker 00

  Absolutely. I mean, I think a lot of founders start companies for the wrong reasons. And I was one of those founders. I was, I remember very clearly, I was at Twilio for three years from intern IPO and then stayed to lead a team. And I'll never forget the day I knew I had to quit. I just had this big architecture decision and my tech lead had given me a bunch of comments on it. I was like, this guy, I disagreed. And I went home and I coincidentally got a text from my future co-founder of my first company, Len. And he said, hey, like, do you want to help me hack on this thing this weekend? I was like, let's do it. And we built this really awesome product called Stream Club. And the reason I started that company was because I liked the building. I liked the idea of starting a company. I wanted to see what I was capable of. But I wasn't like, I didn't have any like differentiated attitude

• 42:36Speaker 00

  towards the live streaming market back then, and nor did any of us really. But we were at the right place, right time and had the right team to go try and build a great company. And I think that people who start companies because they want to see what it's like, it can be very challenging because you don't have that like raw emotional connection to what you're building that will push you through all the hard times. You'll see what it's like. It's really hard, you know? You're building a company is very painful and you lose your hair and you don't work out as much and your life is not the same. You don't see your friends as much. So I think that with after we exited Stream Club and we were able to go work at Mux, which is a video in for company, really great company. I took a lot of time to reflect on why should you start a company? And to me, you should start a company when nobody else is building the thing that you need.

• 43:21Speaker 00

  And with Stream Club, I spent a lot of time on browser automation or this idea of the headless browser, which is what BrowserBase offers is a browser running on a server. And I spent so much time on this stack and I became very opinionated about it. It's one of the reasons why we sold Stream Club to Mux was for headless browser technology. And that really helped me develop a point of view on this infrastructure and why it was important. And maybe a year or so later, this rise of AI and AI agents, I kept getting calls like, hey, how do we use this headless browser stuff? You're super familiar with it, like you know a lot about it. And I just gave them a lot of feedback. I said, oh, well, I'm sorry, guys, like most of the infrastructure here kind of sucks. So you're going to go build it yourself and here's my tips. And by like the 20th conversation, it became very clear that no one was building the infrastructure

• 44:05Speaker 00

  to help these people automate the web and that needed to be me. But I also was like very hesitant to start a company. I felt like to start a company, you have to be willing to commit several years of your life to building something great and also be able to, you know, sacrifice quite a bit. There's a lot of opportunity costs these days and if you're like a well, well talented person. So it was like kind of a last resort thing where it's like, no one's going to build this thing. I know how to do it and I'm probably the best person out there to go build this company. I should do that company. I should start that. And that's really what was the driving force for BrowserBase.

• 44:38Molly O'Shea?

  And you're a solo founder. That's a lot of stress. Startups are a lot of stress. We're really fortunate to be sponsored by Brex and Brex really is like the best partner for startups. And I'm not just saying that, but they also really stress on performance. And so like, how do you keep your mindset clear while you're building, while you're running and doing these like really hard sprints and loops?

• 45:02Speaker 00

  Yeah, well, I'm not a solo founder by choice necessarily. I tried to get like three people to be my co-founder. And they also know. So like, I, I think in the end that worked out for the better. And I do really love being a solo founder. Specifically because as a solo founder, you remove this alignment step within co-founders where you have, we have several co-founders. Often the co-founders have to get aligned and then you get aligned to the rest of the company. So when you're a solo founder, it's just you and the company in constant harmony. And if I feel like the company's not moving alignment to me, I could have an all hands meeting and we could kind of realign. And our all hands meetings kind of feel like mega church vibes these days, you know? It's a lot of like talking about the values and talking about the customer stories. And I think that really helps us stay in tune as a culture.

• 45:47Speaker 00

  In terms of taking care of myself, the, the most important thing is like finding a great partner. And I'm very lucky. I just got married this year, so I'm still in the honeymoon phase.

• 45:54Molly O'Shea?

  Congrats again.

• 45:55Speaker 00

  Yes, thank you. So like, I think as a solo founder, you have to have a person that's going to be there for you in the rest of your life and all the other things you have going on.

• 46:04Speaker 00

  I used to be super fit. I like did an Ironman. I've run three marathons. That has gone out the window. I've never seen a series they founded with a six pack. So I think it's something you have to sacrifice, you know? And I, I, I try and like do my best to, you know, get to the gym where I can. But in the end, my goal with the first year of Brown Space was to be obsessed. Like, I want to be obsessed with a problem and sacrifice everything else to make this thing happen. And I think you have to make a lot of sacrifices to make that work. But if it starts working, the momentum is what fuels you to keep going. And post series B, definitely trying to hire a little bit more, get a little more time back. So I can maybe start to like, eat a salad every once in a while and go for a run.

• 46:46Molly O'Shea?

  Good.

• 46:47Speaker 00

  Working on it, you know?

• 46:48Molly O'Shea?

  Get a protein shake.

• 46:49Speaker 00

  Yeah, get a David Barr, right?

• 46:51Molly O'Shea?

  Yeah, you could. You could. I think David Barr uses Brex.

• 46:55Speaker 00

  Oh, there we go.

• 46:55Molly O'Shea?

  So there we go. Double collab.

• 46:57Speaker 00

  Boom.

• 46:58Molly O'Shea?

  I also, I can't help but notice Paul Klein, you are the fourth Paul Klein. Tell us about Paul Klein.

• 47:05Speaker 00

  Yeah, there's

• 47:06Molly O'Shea?

  All of them.

• 47:07Speaker 00

  This, well, you can think about Paul Klein, uh, Dua Lipa's ex, the singer, actually of Lanny. Or you think of Paul Klein, uh, Ohio family. And my, my grandfather is, is someone I really look up to. He, uh, he started out, you know, it's like going through school, night school and grew to be, you know, this kind of like, you know, business executive, working for a company called National Cash Register and then doing some turnaround exec stuff later on. And, um, I think like, we have it on easy mode these days, starting businesses. Like I can go raise capital. I can go, you know, hire people. I can use the internet. I think it's, was so much harder to build great businesses in like the eighties and nineties. And whenever I send them my investor update, the guy has a very high bar. Recently I invited him to our series B party, um, happening tonight.

• 47:51Speaker 00

  And he was like, hey, this is great. I appreciate it. I'd rather only come to an IPO party. Um, so let me know when that happens and I'll come on by. I was like, oh crap. Okay. Well, better, better hustle up, but, um, I think that's just a great attitude of like, you know, the bar for companies is, is, it's much easier these days. And, um, when you meet people who've been in the industry, and I think actually our board, Glenn Solomon is someone who was really has a lot of industry experience. Those are the people that can really help you connect with like, look at all the stuff you have to overcome in the future. And don't take it for granted that you can go raise capital and you have Brex to help you do your finances. Right? Like the tool stack for modern founders is much simpler now, which means I could spend more time on product and customers and hiring than, you know,

• 48:35Speaker 00

  all the other admin stuff I may have had to do 20 years ago.

• 48:38Molly O'Shea?

  As we wrap up, I love to go through two extra segments, um, on future outlooks. So I'd love to know, what are you most looking forward to this year?

• 48:48Speaker 00

  What I'm most looking forward to this year, I am really looking forward to the things I can't predict, honestly.

• 48:55Molly O'Shea?

  Okay.

• 48:55Speaker 00

  Like there's, there's so much that's going to surprise me. I think there's going to be new model releases and there's going to be like new companies that break out. Like who's going to beat cursor on the path to a hundred million of ARR? I think there is so much surprise that happens in our industry and I, um, I love that surprise. I love waking up to a big release. It's like, oh wow, this is so exciting. So continuing to see the innovation is gonna be pretty fun and I I don't know what's gonna come, but I know it's gonna be really cool and it's gonna push our technology forward quite a bit.

• 49:24Molly O'Shea?

  Love that. Our final segment is with Kalshi, so we'll bring in a Kalshi market or two. And these these will be on the theme of like, you know, AI and software and infrastructure and anything that we can pull. But with AI agents, I want to go back to like, I know, by the way, BrowserBase is not an AI agent company, it's infrastructure. But of the theme, so jobs are like one of the hottest topics. And whether or not people are gonna have their jobs replaced, maybe BrowserBase ends up taking over some people's jobs, I don't know. But how high do you think unemployment will get this year?

• 50:00Speaker 00

  Well, first of all, I think it's not gonna replace, it'll augment. I think there's still gonna need to be humans in the loop supervising. And if anything, I hope that the work that BrowserBase does helps give people more leverage in the work they do. Instead of you doing the boring work, automate the boring work, do the exciting work and like spend more time like doing the boring work faster. So I I don't think it's gonna replace all the jobs. The question is like, how high will employment get?

• 50:23Molly O'Shea?

  Unemployment.

• 50:24Speaker 00

  Unemployment get. Um, I know people have been saying some pretty crazy stats. I'm I'm gonna say a baseline nine percent. I don't know where we're at right now. I think we're probably in the eights.

• 50:35Molly O'Shea?

  Nine is pretty high.

• 50:37Speaker 00

  What I don't know what the baseline of employment is right now. Can I pull over and look at this?

• 50:40Molly O'Shea?

  On Kalshi, it has a forecast of eight point seven, so you're pretty close. Like, you're pretty close. Majority is is shooting for above five percent, which is kind of a cop out because it's pretty easy.

• 50:52Speaker 00

  I saw a stat yesterday at the Red Point Infrared 100 Summit, which we are.

• 50:57Molly O'Shea?

  Oh, how was that?

• 50:58Speaker 00

  It was really fun. And apparently, and this is such a bummer because I'm a CS major, but new grad CS majors have a larger unemployment rate than art history majors.

• 51:07Molly O'Shea?

  Really?

• 51:07Speaker 00

  In 2025, yeah.

• 51:09Molly O'Shea?

  Who would have known?

• 51:10Speaker 00

  I art history making a comeback.

• 51:11Molly O'Shea?

  By the way, I went to college for art, so.

• 51:14Speaker 00

  This is your revenge art.

• 51:15Molly O'Shea?

  I pivoted quite fast, to be fair, but artists are coming to get you.

• 51:20Speaker 00

  Artists, you should learn to art.

• 51:22Molly O'Shea?

  You should learn to art for sure.

• 51:25Speaker 00

  I mean, seeing that, I would, I'd say like, yeah, you want, you want high ROI on your, your gamble, right? So I'd go nine. Let's see what happens.

• 51:32Molly O'Shea?

  For sure. I agree. I like to play the risky game. Um, okay. This one is like kind of spicy, but especially in SF, what do you think the best AI will be this month?

• 51:44Speaker 00

  I think it's Gemini. I think the, the DeepMind Lab has been doing a really fantastic job. I think it's really tough because I always see like these stats of like DeepMind engineers leaving to go to places, but the team really did a great job. You know, we have early access to the Gemini computer use model at browser base and it's very, very good. I'm excited for the longer, the larger release of it in the future. Um, but the Gemini team has really been doing a great job. So I'm going to bet on Gemini all day long. I don't care what the odds are. I'm doubling down. We just raised 40 million bucks. Put it all on Gemini.

• 52:16Molly O'Shea?

  Put it all on Gemini. You're not alone. Uh, most of the CTOs and like engineers and more technical people that I talk to all are hot on Gemini for sure. This one's a curve ball. Are you ready for this?

• 52:29Speaker 00

  I'm ready.

• 52:30Molly O'Shea?

  Will a major meteor strike Earth before 2030?

• 52:34Speaker 00

  Define, what's, define major?

• 52:36Molly O'Shea?

  10 kilotons TNT equivalent or greater?

• 52:40Speaker 00

  That seems kind of big.

• 52:42Molly O'Shea?

  Sounds big.

• 52:42Speaker 00

  I'm gonna, I'm gonna go no. I'm gonna say no way. Though I'm sure there's some data and there's some like space nerds that are saying, actually, we will have one. Uh, but my uninformed, hop, optimistic opinion is no, we won't be struck by a meteor by 2030.

• 52:58Molly O'Shea?

  Okay.

• 52:59Speaker 00

  What, what's the world saying?

• 53:00Molly O'Shea?

  There's a 65% chance.

• 53:01Speaker 00

  Oh geez.

• 53:04Speaker 00

  We'll, we'll get to AGI and then immediately go back to the caveman days when we get hit by a meteor.

• 53:09Molly O'Shea?

  That's a really high chance.

• 53:11Speaker 00

  Yeah, well, you know, maybe we, our AGI will be able to dodge the meteors.

• 53:15Molly O'Shea?

  I would hope the AI can turn or deflect the meteor away from the Earth.

• 53:21Speaker 00

  It's, it sounds, it sounds possible. Maybe if we all just yell at the meteor loud enough, like it'll spin around.

• 53:26Molly O'Shea?

  We should deploy some browser base to hit the meteor.

• 53:31Speaker 00

  Deploy some browser base. Yeah. The web browsers collectively will maybe like make a DDoS attack and point it at the meteor.

• 53:39Molly O'Shea?

  Yeah.

• 53:40Speaker 00

  You know what, if I have to, we'll raise a Series F and we'll ride the rocket out there and we'll face the meteor ourselves.

• 53:45Molly O'Shea?

  Please, we're all hoping, waiting, counting on you.

• 53:48Speaker 00

  Do my best.

• 53:48Molly O'Shea?

  Okay, Paul, it was amazing to have you on. Congrats again on the Series B.

• 53:53Speaker 00

  Yeah.

• 53:54Molly O'Shea?

  I'm so excited for you and this was so fun.

• 53:56Speaker 00

  Thank you. Thanks for having me, Series B. Let's celebrate.

• 54:00Molly O'Shea?

  Hey, it's Molly. If you enjoy our interviews, check out our newsletter, sorcery.vc, where we deliver a once a week top deals and tech headlines email and also go deeper on our podcast interviews. Subscribe to Sorcery today. And don't forget to subscribe to the podcast on YouTube, Spotify, Apple, or wherever you listen. Link in description to sign up.