How to Raise + Scale an AI Startup in 2025 | Agentic EP 07

AI is a massive platform shift. AI startups are exiting much faster than everyone else. Seven years versus 13 years, which is great for returns for VCs and great for outcomes for founders. This interesting analogy to electricity, you might have heard about some of that.

This is most like electricity. I guarantee you there is not a single application that you can think of is not to be made better.

Some people think it's hype or a bubble. We all realize that big changes are coming. AI grab 37% of total funding for the entire year globally. All of these are all time highs. We've never seen anything like it. In the 50 years of venture, I still see a lot of especially first time founders make the mistake of thinking that your job is to put together a deck to go raise money. And that's just not the way it works. The main thing is not funding. The main thing is that you're solving an important problem that could become a venture scale idea. If you're building something at the compute layer, you're building a new AI chip or if you're in the LLM layer and you have a new take on how to do LLMs, probably it's going to be the technical risk. So you need to prove that you could do that. You're in thinking the MER that has the pain. You're going to hear entirely different things. It's going to truly guide you on your journey. It's not going to end up being whatever you have in your vision document or in your PRD or in your mvp. Welcome to the Agentic podcast where we dive deep with the pioneer AI founders reimagining business and investors betting on them. We, we bring you the war stories, insights and perspectives you won't hear anywhere else. I'm your host, George Van Darrien, the general partner of Untapped Ventures, a VC fund that invests in early stage AI startups. Let's go. Hey. Hey and welcome back to the Agentic podcast. Today we're trying something a little bit different. So instead of our usual interview format, I'll be sharing a talk I gave to a group of AI founders in SF called the Founders Bay. Shout out to Marianne Becker for inviting me. And in this session I dive deep into what it takes to build a category defining AI startup in 2025. One that attracts investors, wins over customers and assembles an all star team. I'm going to do a presentation where I share the state of AI 2024 and Q4 funding stats including round sizes, valuations, exits. Then I talk about ideation and how to think about what to build and validation, how to think about what you need to validate to track capital. Next, I cover traction and how to get customers, recruiting and how to build a team, leading up to a discussion about fundraising best practices. And finally, I touch on how we as a firm help our portfolio companies. And then there's some really good Q and A at the end that I'm sure you'll find valuable as well. So if you're a founder just getting started, trying to figure out what to do or gearing up for a fundraise, this episode is packed with no fluff, just practical, actionable insights from the trenches designed to set you apart. Have a listen and let me know what you think. Let's go. Glad to be with you all today. So, again, my name is George Van Darien with Untapped Ventures and I think you're in for a really exciting discussion today. So we're going to talk about how to build a native AI startup that attracts funding, customers and talent. We have a pretty meaty agenda, so I'll provide a brief intro and then we'll dive into the state of AI, including all the latest Q4 and 2024 funding stats, valuations, all that good stuff. The numbers really show why AI is still the best industry to build in. I'll share some innovations, videos, demos, and then we'll kind of walk through some our perspectives on how to think about ideation, validation, traction, recruiting, fundraising, and then how we add value to our portfolio companies. I'll offer you guys a gift at the end and then open it up to Q and A. Does that sound good? All right, awesome. So we're going to flow quickly. It's probably about 70 slides, a lot of good meaty content. So just a little context on me. So I was born and raised here in la. Again like we said early on, thankfully not affected by the fires, but that was definitely sad situation of what's going on. I went to USC, graduated from there, and then my passion with AI began in my 20s after I read this book, the Age of Spiritual Machines by Ray Kurzweil. And I was fascinated by this idea that we're going to build computers that would feel like they have sentience. And here we are 25 years later and we've done that. I went on to build a company called ami. We helped companies eliminate paper and automate manual processes with digital workflows and automation and AI using computer vision and ocr. And by the end we were knee deep in workflow and bpa, BPM rpa, all the kind of workflow automation by the time I exited. And then I started angel and LP investing and doing soul searching to see how I wanted to discover my life purpose and how I wanted to spend the coming decades. And as a big personal development and Tony Robbins guy, I wanted to help humanity reach our untapped human potential, hence the name Untapped Ventures. And for us to become superhuman by today's standard by being able to leverage these tools, specifically fostering innovations and breakthroughs in AI. So a couple years ago I launched Untapped Ventures Fund 1 to invest in early stage AI startups that, that uplift humanity. We're off to a great start. We have, you know, we raised a bunch of money. We have 31 investments, a handful of markups and couple exits. We write 500k checks into Precede and seed stage native AI startups. And so that's a little bit about me and the fund. So now that you have a little bit of context, let's dive into the state of AI. I think that most everybody on this call realizes that AI is a massive platform shift, right? And it's going to impact just about every aspect of our personal and professional lives in the coming decade. The fabric is going to be woven into everything. And, and I think we realize that this is not something to confuse with other technologies that didn't quite live up to their full potential. And there was more fad and hype involved. You know, if you look at Goldman, Goldman Sachs says The investment in AI could reach 200 billion this year. And then Zion says it could double to 422 billion by 2028. And then Goldman points out this interesting, that 2% of GDP or more, but this interesting analogy to electricity, right? And if you're following AI, you might have heard about some of that. So like I said, I'm going to play a number of videos and demos as we go through. So you know, on the topic of this analogy to electricity, let's hear from Jeff Bezos.

What is it that you're doing at Amazon?

AI.

AI, yeah. So I mean there are a few things, but small, but it's 95% AI.

And where do you.

Because we are literally working on a thousand applications internally. So you have to remember AI, modern AI is a horizontal enabling layer. It can be used to improve everything. It will be in everything. This is most like electricity, these kind of horizontal layers like electricity and compute and now artificial intelligence. They, they go everywhere. I guarantee you there is not a single application that you can think of that is not going to be made better by AI.

So I love that, right? He's come back in and he's 95% focused on AI and this analog again to electricity, right? So there's, you could look at all these people in the photo and they had no idea what was around the corner for them and how drastically their lives were going to change. Going from using product and services like candles and oil lamps and fireplaces for light and heat and washing clothes by hand and you know, using manual tools like a hammer, to how all these power tools and you know, electric devices inside the, the home and the business were right around the corner. And I think that's the way that most of society is today. They hear the headlines, some people think it's hype or a bubble. And I think we all realize that big changes are coming. So it's no surprise to see that AI got 20% of all global venture funding last year, right? Or sorry, in Q1, in Q2 that doubled to about 28%. The dollars double to and the percentage went up to 28% as you could see on the chart on the right. And 35% by another count. That's, that was the highest ever recorded. And then in Q3, while the overall deal count and deal total deals, venture dollars went down on the left, you could see that on the right, I took a larger share, 31% of total total global venture funding. And so that continued in Q4. Even there was a bounce back in total venture dollars globally. A lot of that was AI. I mean, AI went from 17 billion in Q3 to 44 billion in Q4. And so basically AI is carrying the entire venture market or also every other category is struggling. And AI is the bell of the ball, right? The shining star. So we definitely see that broader Trend continuing into 2025. If we look at the total numbers, you could see 55% increase in funding to AI. And overall AI grabbed 37% of total funding for the entire year globally, which is just a staggering stat. All of these are all time highs. We've never seen anything like it in the 50 years of venture. If you look at it, a big part of that was the ultimate mega deals, right? I don't know what to call it. Mega deals used to be a hundred million dollar rounds and above. These are multi billion dollar deals. And so that's a big part of it. But if we kind of come to where a lot of you guys are at in the early stages, we could see here what the valuation premiums are, right? So we could see that AI startups are commanding a higher valuation. There's no precede here seed 17.8 million versus 12.8 for non AI counterparts. You know, I think for pre seed a lot of what we see is, you know, it's in that 8 to 10 million, 5 to 10 million kind of dollar range. And it's no surprise that AI is creating the most unicorns, right? I think 32 out of 71 unicorns in 2024, 45% of all unicorns. And now we're getting some early exit data that AI startups are exiting much faster than everyone else. Seven years versus 13 years, which is great for returns for VCs and great for outcomes for your pocket for founders. And then if we look at the AI agents or if we look at what's driving it, the AI agents are really leading the way, right? So AI agents as well as kind of customer support operations. Although I think that's been filled out from an early stage investing standpoint. So 2025 predictions. You know, we think that AI adoption is going to accelerate in the, in the enterprise environment. We're going to be seeing the rise of all kinds of agents, including code agents, back office automation. And the key is going to be the TTF fee, right? The time to first value. How quickly can you deliver value to your customers? If they can get up and running, you know, within minutes or hours, that's going to be significant advantage. So all right, so that gives us a quick lay of the land and we can go a little bit broader. But now some of you already know what you're building and you're well on your way. Some of you are a little bit earlier on, but just to help continue to set the horizon and paint the picture of what's happening. You know, I think there's a lot of exciting innovations that are happening in and around generative AI. AI is being used in drug discovery and design. It's being used in materials sciences to create new materials in neurotech and synthetic data, robotics, quantum AI. And I could go into more detail details on all these. But just kind of based on our initial feedback, I think I have a sense of what you guys are looking for. I will play the figure robot video because I think it's so cool. You know, we're really excited about the robotic space and humanoid robots coming onto the scene. Here's the new figure 2. If you haven't seen it in live in a plant at the BMW plant in South Carolina.

Sam.

I love how he puts it wrong and then he just kind of like jams it into place. So that's going to be really exciting. I'm also very excited in what's happening in quantum. Google had a big quantum announcement in December which was kind of mind blowing. I'll, I'll let you have a listen.

That we developed a new very high quality quantum chip and this enabled two breakthrough computations. And one is an error correction where we showed for the first time that as we use more qubits, the error rate comes down and it comes down exponentially. The second important achievement, breakthrough achievement, is that we did benchmark computation. Took just five minutes on our new chip that would take the top supercomputer 10 septillion years to perform. So 10 septillion audience may not know that's one with 25 zeros. So it's a mind boggling long time.

10 septillion years, that's a mind boggling number. And they did it in five minutes. Pretty, pretty amazing stuff. So the area of AI that we're most excited about is agentic AI, right. I think this is what's really going to be transformative, especially in the business context. So if we just take a step back and kind of look at the generative AI timeline, right. This is a simplistic view. First we needed compute. Nvidia lucked into it, accidentally stumbled into it, right. They had these GPU chips for graphics and gaming, but those are the best things and they could make LLMs possible. That allowed for OpenAI and Gemini and others. Then this is a thing. So now we need infrastructure. So you have people like LangChain building infrastructure. And then from there people started using the OpenAI APIs to build horizontal solutions which were initially called Chat GPT wrappers. Right. To a certain extent, you know, doing copywriting and marketing and emails like Jasper. And then from there you had the deep domain experts like some of you coming out and creating vertical AI solutions. So like Harvey did for legal. And now we're in the era of agentic AI on our way to AGI. So hopefully everybody here is familiar with what agentic AI is. It's the next big thing. That's what we're aggressively investing in. Agentic AI is a kind of AI that can act independently to achieve goals and complete tasks with minimal intervention.

Right.

It's this idea that, you know, we're moving from search in Google to we've had answers in chat GPT the last couple years. Like I don't want to search and look through 20 blue links, I just want the answer to something. And then now we're moving into action, right? Agents that are going to do the things that I, I need to have done. So, you know, I think it's helpful to again see it in action. For those of you that aren't familiar, here's just an example of one agentic solution.

I'm here to introduce Jace, your AI coworker. With just one instruction, Jace can handle any task for you. For example, Jace can plan a full fledged trip. It will ask you for all the necessary details and make all reservations for you. This is possible thanks to our autonomous WebAgent 1 model, AWAONE. With AWAONE, Jace can control and perform actions in a browser just like a human would. Let's take a look at the first few steps in the Live View, Jace is starting on Google, going to the Airbnb website, entering all the details and clicking search. Jace isn't just an assistant, it can take on entire roles. Imagine a recruiter who never sleeps. Jace finds candidates, schedules, interviews and manages the hiring process effortlessly.

So that's a good demo. One of the examples. So like when you see a slide like this, this is one of our portfolio companies. So Harper built an Agentic AI solution for insurance, right. And it's really exciting. It uses phone, text and email to get a hold of the person and then there's a voice AI. And, and what, what it's able to do is effectively sell any consumer or business certain insurance ENS insurance. And what's interesting about agentic is they are not pricing based on usage or calls or seats. Right now you have the ability to price based on outcomes. And so they were pricing based on a percentage of the policy that they sold for that insurance agency and then the insurance carrier. Right. Like State Farm for example. But what they quickly figured out is they could just do this for themselves and build a fast growing, highly profitable, probably faster growing and more profitable than any other insurance company if they just did it for themselves. So that's the go to market that they're pursuing is becoming an insurance agency themselves and then being able to use their agentic solutions and use agentic all throughout, not just in sales, but in claims and every other facet of the business. So we're really, we're really excited. We, we caught them early on. They're, they're in the winter 25 cohort of Y Combinator. We have a couple of portfolio companies that are in the current Y Combinator batch. And so with the gentic, what we have is this idea that we're moving from, you know, you're the human controlling the chatbot and the AI is your co pilot like Microsoft branded to really the AI can become the pilot and we have humans in the loop just making sure the process is running right. And really where we're moving to is this idea where for many business processes we won't need a human in the loop.

Right.

It could just run automatically. Just like many API calls and integrations run today and it's normal there is no human things just run. It's not that far fetched if you think about it. We've already done a very extreme use case of this which is the Waymo cars, right. A lot of you guys are in San Francisco. I mean you know that's a great example of an agentix solution in action. You know, we've taken something that's very dangerous and very important as it involves human lives and we've been able to get it where it can continuously do the job. Taking me from point A to point B without a human driver.

Right.

And I think we're going to see this spill into many different business use cases, accounting, hr, finance, healthcare.

Right.

Coding and some of the areas that you see on the screen. So you know, I think it's a really exciting time with agentic. I, I love this Steve Jobs connect the dots speech. So I'm sure you guys are all familiar with it but it's so good that it's probably worth listening to one more time. And the reason I'm gonna play it is because I always wondered what, what my connect the dots moment would be.

It was impossible to connect the dots looking forward when I was in college, but it was very, very clear looking backwards 10 years later. Again, you can't connect the dots looking forward. You can only connect them looking backwards. So you have to trust that the dots will somehow connect in your future. You have to trust in something, your gut, destiny, life, karma, whatever. Because believing that the dots will connect down the road will give you the confidence to follow your heart even when it looks lead you off the well worn path and that will make all the difference.

So you know, I hope that was a little burst of inspiration as you, as you reflect on maybe what your connect the dots moment is. For me it's right now, right because in that prior company that I mentioned we were doing workflow automation for you know, over 15 years and, and I worked with every vertical and horizontal you could put on a chart like this. Military, government, education, banking, transportation. As far as the vertical as, as the horizontals we did HR automation, AP automation, billing automation. We, we did it all and we integrated with ERPs and CRMs. And so to me, Agentic AI is just, you know, workflow 5.0 and it feels very natural to me. It's very easy. And, and all those years that I put in, I feel like that's my connect the dots, that now it's really fun for me to meet with Agentic AI founders, be able to quickly understand what markets they're going after. Is this, is this a good investment? And then more importantly, how do we help you once we make an investment, right. How do we help you with go to market, pricing, packaging, marketing, sales, strategy, storytelling, fundraising and all the areas that is is our superpower and my superpower. So we're doing a lot of activity around this. You know, we had a series of events a few months ago where we had some top tier VC funds and pure play funds like Khosla. And we had six Agentic founders. We did that in SF for SF Tech Week. That was a great event. You know, we had over a couple hundred people at that event. You know, we did something similar the following week at LA Tech Week, another great event. And my alma mater at usc, we launched a podcast last month called Agentic. We're off to a great start. We just released episode six today and on that episode is the tat. Sorry, that was their former name, Harper Insurance that I mentioned earlier. That was a great discussion with the founders. So, you know, we're obviously very bullish on this and if anybody's building in that space, definitely would love to hear what you're up to. And I think through those events that we did, there was some consistent feedback, right, that you know, from the founders that, you know, three kind of key takeaways, you know, feedback loop is critical in terms of like having that in your Agentix solution. You know, balancing the fact that outcomes are the top priority and time to value like we talked about. And similar but different feedback from the VCs on our panels that might be helpful to use is that there's a lot of training data that's untapped in the enterprises.

Right.

Everybody's tapped the public web, Reddit and everything else that's running out. But every company has its own unique proprietary data set that's dark data, not bad data, but dark data, unorganized, underused. So figuring out how to tap into that is good use cases and really developing your key edge. If you have something that is like a chat GPT wrapper. Okay. I mean a Lot of sasses were just databases with the SAS wrapper. Right. So what really matters is, is what's the moat, what's the defensibility, what kind of unfair advantage do you have to be able to win in that market? Another thing that we spent quite a bit of time on in all those discussions and even when I talk to guests on my podcast, is this horizontal versus vertical AI or agentic AI debate. And it's really interesting and my takeaway is that both can win, but they each have different playbooks and they make sense to start at different times. So for example, you know, if someone on the line is wanting to start a customer service agentix solution, I wouldn't recommend that because you already have Maven AGI, you have Sierra, Sierra's Meta's prior cto. He's raised a bunch of money. Maven's raised a bunch of money. I had them on the podcast, they were my first episode. You know, you're a little too late on that, right? For a horizontal customer support agentic solution, it's going to be difficult to prove that you're going to be able to win in that market versus these people that are out to a running start. But a vertical solution, that is an entire post sale, you know, support system for a specific vertical and you're going to do that better than Maven and Sierra. That could be interesting. As long as there's a big enough tam, you know, if you say you're doing something for, you know, it consultants. Okay, great. What's the size of that market? How do you go into other markets? So there's two different playbooks and the timing is really important. So, so hopefully this is all helpful. We're going to get to the other section soon, but that's what we look at, right? And I think that's what you want to think about is, you know, what's your moat, how mission critical is it? You know, and VCs like us, we have a ton of different diligence questions that we'll ask and all that good stuff. But once you've come up with the idea and I tried to kind of paint the landscape and again just take, you know, try to talk to everybody in the room. Some people are still figuring it out, some people have an idea and they want to validate it, which is what we'll talk about next. And then, you know, some of you are further along and you're working on traction and fundraising. So we'll get to that. Now this is part of a larger, longer presentation. So I'M not going to be able to go through everything. But the short version of validation really boils down to, you know, what do you need to prove to de risk the business and raise your rounds? You know, I think I still see a lot of, especially first time founders make the mistake of thinking that your job is to put together a deck to go raise money to go do the thing you want to do. And that's just not the way it works, not even in AI. Your job is to go prove that you can do the thing you want to do and then come to VCs and say, hey, look, I proved it, now I want your money, right? So you have to figure out what do you need to prove now in the four levels of value creation. If you're building something at the compute layer and we have a few investments in that area, if you're building a new AI chip, right, or if you're in the LLM layer and we have a few investments there as well, and you have a new take on how to do LLMs that, that doesn't cost as much or as faster or is better, right? Probably it's going to be the technical risk, right? So you need to prove that you could do that. But if you're in the infrastructure or the application layer, we're not as concerned that can you build it. You know, as long as you're technical founders and we wouldn't invest in anything that doesn't have a technical founder, but you know, as long as you know your technical founders and you know how to build it, that really, it becomes a market risk, customer risk, finance risk and all that. So you need to, you need to de risk those before you bring it even to a pre seed VC like us, right? Because the bar after the, you know, the collapse that we had a few years ago when VC stopped drinking and doing cocaine and got back to VC basics, with the exception of a little bit of hype and AI is back to like, you know, prove that you have something worth investing in and then we'll invest. And so I think that's what you need to figure out. Now assuming you figured that out, you're doing discovery and for a lot of you guys, I think based on what I saw, you're building typical application layer businesses. So it really is the customer risks, right? So you know, have you talked to 100 customers? What did you learn? How many, you know, how many customers did you get out of that? How many pre sales, early orders or design partners?

Right.

In the old days we used to call Them, you know, beta clients or founding members, but now we all call them design partners, right? So get your design partners, whether they're paid or unpaid, people think that it has to be unpaid. You know, I think your design partners could pay if the problem is big enough, they could pay and then leverage AI tools, right? So don't. In addition to the AI you're building and in addition to ChatGPT, which is obvious, but you know, make sure you're also using all the other, not all the other tools, but that you're picking from the other tools and using what is going to be beneficial to you.

Right?

So just as an example, for fun, so, so here is my. This is one tool that we use. This is. You're probably familiar with heygen. So this is my AI avatar. Again, I, I didn't record this video. This is my AI avatar. Hey everyone, if you're hearing this, that means we're at the point in the state of AI venture landscape and investment opportunities webinar where George is showing off his magic tricks, including this AI avatar. Pretty amazing what's possible with generative AI right now, right? So are you going to go all in and commit to AI personally? Will you take it back to your teams and adopt the latest tools? And most importantly, are you ready to jump in and make smart investments in innovative AI startups? And then the beauty about a tool like that is if I want to go after different markets now, this could scale our video production, right? But let's say I want to go after different markets. I could select it out of the box to 25 different languages, right? So let's say if I'm, you know, we want to expand and we're now looking for AI startups in Latin America. And while I'm born and raised in LA and I speak a little Spanish, it's not perfect, right? So my team could easily flip the script and it's, it's not a prompt, it's a script. And we could, they could flip it to Spanish and it sounds something like this. Any native Spanish speakers? I see a Jorge here.

Pretty good, right?

What was the tool for the AI avatar? That was heygen. H E Y G E N All right, so now let's talk about recruiting.

Right?

And again, I'm going to give you the short version, kind of like what I said on the validation front. So on the topic of recruiting, there's a lot of great strategies for how you're going to recruit at a high level. What I want to see, what people like me want to see is that you are a visionary. You see a future that does not yet exist and you're able to convey that to people and you're able to get them to jump onto your journey. You know that you're able to get like top notch, gainfully employed engineers in San Francisco or wherever to leave their great jobs at Meta or OpenAI or whatever and come join you on your journey. Not go hire a dev shop in India or wherever, right? But actually get technical people, including founding engineers, to come join your journey. Your co founders, advisors, customers and investors. So, you know, recruiting is really all about being able to sell the vision, being clear in how you articulate it and being able to get people on board. So happy to take questions on that, but I'll just kind of, for time's sake, leave it there. And now let's go ahead and switch into fundraising. And I've already been talking a lot about the key points, but let's, now that we have a baseline understanding, let's kind of drill down into it.

Right?

So there's two kinds of companies, right? There's people that are using AI in their SaaS and all that. And then there's native AI.

Right.

And when I say native AI, in my mind, what we're talking about here is we have technical founders that have a master's or PhD in computer science or AI. There's usually two or more co founders and they're building along those four levels of value creation that I talked about.

Right?

You have the compute, the models, the infra and the applications.

Right.

And then as I said earlier, you know, my, the way we see it from an investment standpoint is that for early stage investments, it's going to be hard in any of these.

Right.

We're skating to where the puck is always going to go.

Right.

And so I think the vertical AI is probably coming to an end soon as there's been a lot of proliferation and we're really in agentic AI on the way to whatever's next. Quantum Robotics on the way to AGI. So that's what we're actively looking for. Of course, we've made a handful of investments along the way in those categories, but going forward, that's kind of where we're at. You know, we invested in Xai. I think Elon is the absolute goat tech founder, right. If Tom Brady is the goat in football, Jordan's the goat in basketball, and Elon's the goat when it comes to a, you know, venture backed tech startup founder, right. I name, I have three boys I named the middle one's middle name after him. So, so really excited about what he's doing. Science IO was another one of our investments. This was the first exit in our portfolio. They had built an LLM for healthcare to take unstructured healthcare data and make it structured and actionable. So that one was a nice exit. Six months after we invested for 140 million dollar all cash deal. So that's what we're all about. Primarily precede and seed us based or playing a us game. Agentic Vertical Voice Robotics, Quantum AGI. That's what we're actively looking for. And when we find that the value add, I think you kind of have a sense I, I shared it earlier on but it's everything that's going to be outward facing.

Right.

So we're going to help with customer acquisition, investor acquisition and talent acquisition. We have a mastermind that we run with our portfolio of founders that gets together every month. We provide one on one support. We have our structured check in calls weekly and monthly and then we have what we call our Core 100 which includes our are close to 100 investors in the fund who are all entrepreneurs, most of them with exits and retired executives from Deloitte and McKinsey and Oracle and stuff like that. So all these people have made money and now they're excited about what we're doing and they're excited to help you guys our portfolio companies and you know, our founders love us. We have some great testimonials and like I said, we help help with all the little stuff. Right. So we help you raise money. Just as an example, like we were involved with a pitch deck review and you know these might seem small unless you read the details, but you could see the before after of some of the slides here and just knowing what I'm looking for and what others are looking for and making sure that as you're raising the rest of this round or your next round that we get you ready and then of course making introductions along the way. We have a pretty deep network of tier ones of pure play AI VCs and so we have our partners that we bring into the mix. So I've done pretty good on time. Let me go ahead and start to land the plane here. So takeaways and next steps. So hopefully this was valuable. Hopefully you've already taken the red pill and entered the AI matrix and you know you're listening to all this and you're feeling pretty good about the decisions or I've given you some things to reflect on and Consider, you know, go out and build your AI team.

Right.

The A team was one of my favorite shows growing up. So we have the AI team and then there's a gold rush. There's a lot of more money pouring into AI and you know, do the things that are going to help you to capture that, that funding. So with that said, hopefully this was helpful. I'm happy to open it up to questions. If you want to grab this deck and our fundraising playbook, I'm going to go ahead and share a link and if you don't need that and you feel like hey, based on everything I heard, like we're exactly what George is looking for, then go ahead and pitch us. Everything starts from our pitch us application which is on the screen there. You could go ahead and submit the pitches form and that goes into my team and that's how our entire process starts. So I would prefer you not email us unless like you are not actively building a startup and you know we are hiring right now. If you're looking to work at a VC firm and have relevant experience and like any non investment questions you could certainly email us and my team will take a look at that. But otherwise for funding everything starts with the pitch form. So I'm going to drop that, I'm going to drop the gift.

Grab your gift.

Okay. So with that I'd love to open it up to questions. Hopefully that was helpful. Hey Josh, really awesome discussion. As a first time founder coming with a strong AI background in big tech and all the other logos, I'm wondering how should I think about a lead investor? It's when I talk to multiple vcs they are super excited but how do I get the lead to be first join the journey and then formula the round? Yeah, it's a great question. There really isn't a lead and precede. What's considered a lead is the the VC that leads your priced round and does the diligence and sets the price and terms for your round and typically that doesn't happen at a pre seed. Typically on a pre seed you're raising you know, half a million to 2 million on a safe.

Right.

Sometimes people confuse that and they called the person with the largest check the lead. But that's not your lead, that's just the largest investor. Now if you're asking about how do I get a, you know, big check, an institutional VC that could then be leveraged to go get all the other checks. Well it's basically everything we just talked about, right? You have to have everything that would make Yourself attractive, including the fact that you have a well balanced co founding team. That's technical. You have the initial risky points of the business, whether technical or customer or market. You have some basic MVP or something built that to show that you're able to, to do what you want to do. Ideally, you've talked to 50 to 100 customers and you have five to 10 people that are interested to be a design partner. That's all a pretty good start. To be able to, you know, easily raise a, you know, a million on a 10 million post, for example.

Thank you for this awesome presentation. So my question is, so we have built a product, we have sold around 300k. So we have 300k in revenue, but we haven't proven a channel. In fact, we want to change the channel. We haven't raised any money. Are we too slow in raising the money?

Like, no. I mean, especially if you're saying you're changing the channel.

Right.

So I, you know, I think it makes sense. It sounds like you invalidated the channel you had and now you need to go validate the new one. So validate the new one and when you feel confident, then come out the market with that.

And generally like I see it in a lot of these, a lot of these events, like 250k is enough revenue to validate a channel. What kind of revenue is good enough revenue to validate a channel? When do you know that you have validated it?

250K is enough. I mean, the way the way I look at validation is one is a throwaway, doesn't count for anything. Two could be a fluke.

Right?

And I don't know if one of them was your brother or your cousin or whatever.

Right?

But if we're talking about three enterprise customers now, I'm starting to see repeatability. So. Okay, and assuming it was the same, same thing, right? Same problem, solving the same problem with the same solution with the same price point. So the same go to market strategy or channel as you were calling it. So three for me is the absolute bare minimum. And then from there it goes to five is even better. I feel very good about that. And 10, it's like, oh wow, this is amazing. Like clearly there is a, there is a demand that's on the B2B side. On the B2C side the numbers would be bigger.

Right.

So B2C, I would change that, you know, 3, 5, 10 to like 10, 50, 100 at a minimum.

Right.

If it's a B2C solution, and I'm talking about paying customers so, but I'd be curious, like what you learned about your channel that wasn't working and then what you're thinking about moving to. I think those insights are just as important. I don't think we have time for that here. But, but yeah, I think you're thinking about it the same way. And you know, the, the founders sometimes mix up what the main thing is.

Right.

You got to keep the main thing. The main thing, the main thing is not funding. Everybody thinks like I'm going to get customers so I can get funding. Right. And then I'm gonna. Everything is around the next round of funding and that's not the case. The main thing is to, to prove that you have that you're solving an important problem that could become a venture scale idea. What does venture scale mean? A billion dollar valuation. How do you get there? $100 million of ARR times a 10x ARR.

Right.

So you need to be able to show that you are capable of becoming a 100 million dollar company CEO that could eventually go public, that you're going to be a public company CEO, that you have that in you and that you're doing all the right things from the beginning to be able to be on that trajectory. Yeah. So in the early days I think it's, it's figuring that out and if you're invalidating something, invalidate it quickly and move on to the next thing. Your goal is not fundraising. Your goal is building a great business. And the way you know you're on track with that is by getting those initial set of customers that show that there is a large TAM to go. After you do that, the funding will come. You guys have probably heard this in Marianne Sessions, right? I'm sure I'm not the first VC that says that, but you know, I think it's a helpful reminder, as always. Hello, George. And my question to you will be what the verticals you think are less crowded with AI and AI agents for 2025? Yeah, because just almost in every vertical you can find like dozens and already hundreds of companies operating, like I said, definitely vertical over horizontal. I think a lot of the horizontal agentics are already too late for someone that's brand new starting out. And then I think in verticals it just really goes into, you know, who is the team, what is their unique insight, what is their thesis and, and what are they tackling? I think there's a lot of, a lot of areas to tackle, but I think that's going to fill up quickly too. I think there's probably Another year and then most verticals will have agentic solutions that have investments at the early stage like for what we're looking for. And then it'll just be like who's going to win? So you know, I really think it that just 2025 is the year of vertical agentic AI and the time is now. Feel free to reach out. I would say just go ahead and fill out the pitch us form and then we could take it from there. One thing we haven't talked about this entire time with the whole audience is, you know, competition and spending that time up front, you know, looking at the competitive landscape and just being clear about, you know, who's doing what and how much money they've raised, what talent they have and you know, how you're ultimately going to be able to beat them and if, you know, also helping to validate whether the idea you have is worth pursuing or not. I think in every example that I've heard of, you know, there's always, there's always people doing the same thing. And so you know, I think that's another key element. I'm a big believer in being clear about that. I know there are people using AI to help with, you know, synthetic data for survey feedback and lots of those things. So and then there's always the incumbents and the incumbents are, some are sleeping at the wheel, some are innovating, some are doing their best at innovating, but they're all doing something right. So all the incumbent market research, market insights providers are working on some kind of AIs to look at the data, AI for the people, AI for sourcing. So you know, I'd be curious. So like, you know, I think, you know, that was, that would typically be the next thing that we look at is to say okay, this sounds good, you know, you're doing everything the right way, but who else is playing in the space chat GPT and others have made our job easier to do diligence on your competition.

Right.

This is a lot easier to take a look and be able to quickly get a market assessment on your market and be able to see where things are at. There are certain things that are free and doing a deep dive competitive assessment and being really brutally honest with yourself about who's out there, who's doing what and why anybody would want to work with you over them is free and is, you know, invaluable to do early on.

George, I've got a question, quick question for you. Really appreciate the session. I learned a lot. I'm a First time founder. So it's my first rodeo. I am working on a. It's vertical, it's agentic. So I think I've checked both those boxes. It's B2B. But I'm very, very early. I'm kind of in the ideation stage and I've written up my vision document, but I need to translate that to a set of workflows and then translate that to a product requirements definition document and then start working on my mvp. The question that I have for you is you mentioned something about, you know, very important to do a deep dive competitive assessment. I have been in contact with some market research firms. Would you advise going down that route where you hire a professional market research firm to do that competitive analysis or are there other ways of doing it?

Great question.

Do you come from a corporate background? Yes.

Yeah.

So that, that was my sneaking suspicion. So one of the challenges is when founders come from corporate backgrounds, it's common for them to bring a lot of the corporate mentality. A lot of the corporate mentality is a more spread out, slower moving, slower decision making, slower execution velocity. And in startup land, it's exactly the opposite.

Right.

So I would not recommend that you even think like that. Right. So forget about tactic. Let me advise you on mindset, right? You have to have a mindset shift, right. And your mindset has to be. And the, the Harper episode that published on the podcast, that episode went live today. And you know, there was a good segment in the middle where they talked about that. So I think you might, you might enjoy that. I think the short version of it is how did I pick up on that? Because you said you have your vision document and then you're going to turn it into a prd and then you're going to do this and then you're going to do that, you're going to do that, you're going to do that and then you want to hire research, right? So all of those like to me sound like red flags, right? You're in thinking and planning as opposed to like doing stuff, getting out there, talking to customers, creating. There's AI tools that let you create a prototype. When, when I had the idea I found an AI tool, I created a prototype. I started sharing with like customers, investors to get feedback and iterating like in two hours I created a, like a design clickable prototype. I started sharing it with people to get feedback and to guide my, my thinking. I'm trying to make it as valuable as possible. I've been where you're, you're at I've had highs and lows and for me, what's most valuable is when a mentor can just tell me what it is so I could do what I need to do and most importantly, when a mentor can, can help me with the unknown. Unknowns.

Right.

Because there's like, you know what you know that you know what you know that you don't know. And then what kills you is what you don't know that you don't know.

Right.

And so, so, yeah, so, so my goal is to try to be as valuable. If I rub somebody the wrong way, that's okay with me. As long as I've served the other people that were there to learn, I'm okay with that. So hopefully, hopefully you're okay with, with the coaching, life coaching on that. So I could be wrong in my assessments, but I would say stop everything you're doing and move towards talking to customers and validating your idea. I think you already know it well enough and I think just move into the validation phase. Sounds good.

Good advice.

Yeah, I, I mentioned that there's a number of things you could do early on that are free and I mentioned competitive research. But way more important than that is customer discovery.

Right.

It's free. All you have to do is just go out, talk to people. If you're a B2C, go to the mall, go to the beach, go to wherever your customers are and, and go talk to them. If it's B2B, use LinkedIn, use different outreach tools, you know, use interesting messages, get them on, on the phone, you know, get an email, get a survey, do a research report. There's a lot of hacks and best practices that there are that let you get to the people and you need, you just need to talk to them until you're getting to customers. Sometimes it takes time. You could, you know, you start validating it with. Let me rephrase that. You start getting feedback from non customers. The thing I'd be careful about is most of the time it's not that it's not 100% of the time, it's not validation. If it's not coming from a customer, it doesn't count for validation. It puts zero points on the board for validation. But there's still some value and it's helpful in the sense that, you know, you talk to some investors, they give you feedback, they point you in a few directions, they ask you some good questions, you talk to your friend, your old colleague, you talk to this developer, you talk to that developer, you talk to this Other kind of like, kind of customer, but not really because they're in an adjacent industry. They give you their thoughts. All that is great. But the big problem that happens with again founders, first time founders, is that they, they mistake the feedback and the, that noise for signal.

Right.

Because if you go tell like your 20 best friends and colleagues and people that you know they're going to be obligated to give you good feedback and pat you on the back and support your journey. And so it's actually going to mislead you because you're going to be like, oh, George told me to go validate it. I talked to 20 people and I validated it. But in reality that's not what happened. And what happens is when you talk to actual customer that's going to, that has the pain and is going to be paying for it. You're going to hear entirely different things. It's going to, to truly guide you on your journey and your solution is going to morph. It's not going to end up being whatever you have in your vision document or in your PRD or in your mvp. So I, I love being at the place where you don't have an MVP built because when you have your MVP built, you kind of become emotionally attached to it. So it's a beautiful time to spend that time before you build anything to talk to customers and be able to clarify that people have the problem you're wanting to solve, that it's a hair on fire problem. They have budget for it, they're willing to pay for it and all they say is, look, if you can come back and show me you could do this. Yeah, like we really want to be a design partner. Like we want to be, this is so important. We see what you see, we see the future you see and we want to be part of it. And we're okay taking the risk that you, you haven't done this before and that you don't have a product yet. But we'll come along for the journey and see if you could create this for us because it's that important of a problem. Makes sense. All right, awesome. Hey, best of luck to you guys. Thanks for the great questions. Good luck and hope to see you guys in our system and have a chance to talk to you. Bye. Bye. The idea and I tried to kind of paint the landscape and again, just take, you know, try to talk to everybody in the room. Some people are still figuring it out. Some people have an idea and they want to validate it, which is what we'll Talk about next. And then, you know, some of you are further along and you're working on traction and fundraising, so we'll get to that. Now this is part of a larger, longer presentation, so I'm not going to be able to go through everything. But the short version of validation really boils down to, you know, what do you need to prove to de risk the business and raise your round? You know, I think I still see a lot of, of especially first time founders make the mistake of thinking that your job is to put together a deck to go raise money to go do the thing you want to do. And that's just not the way it works, not even in AI. Your job is to go prove that you can do the thing you want to do and then come to VCs and say, hey look, I proved it, now I want your money, right? So you have to figure out what do you need to prove. Now in the four levels of value creation, if you're building something at the compute layer and we have a few investments in that area, if you're building a new AI chip, right, or if you're in the LLM layer and we have a few investments there as well, and you have a new take on how to do LLMs that that doesn't cost as much or as faster or is better, right. Probably it's going to be the technical risk, right? So you need to prove that you could do that, that. But if you're in the infrastructure or the application layer, it we're not as concerned that can you build it, you know, as long as your technical founders and we wouldn't invest in anything that doesn't have a technical founder, but you know, as long as you know your technical founders and you know how to build it, that really, it becomes a market risk, customer risk, finance risk and all that. So you need to, you need to de risk those before you bring it even to a pre seed VC like us, right? Because the bar after the, you know, the collapse that we had a few years ago when VC stopped drinking and doing cocaine and got back to VC basics with the exception of a little bit of hype and AI is back to like, you know, prove that you have something worth investing in and then we'll invest. And so I think that's what you need to figure out. Now, assuming you figured that out, you're doing discovery and for a lot of you guys, I think based on what I saw, you're building typical application layer businesses. So it really is the customer risks, right? So you know, have you talked to a hundred customers, what did you learn? How many, you know, how many customers did you get out of that? How many pre sales, early orders or design partners?

Right.

In the old days, we used to call them, you know, beta clients or founding members, but now we all call them design partners, right? So get your design partners. Whether they're paid or unpaid, people think that it has to be unpaid. You know, I think your design partners could pay. If the problem is big enough, they could pay.