2025 EES Base10: Investing and Engineering: Perspectives on the Evolving Landscape of Applied AI

Hi everyone, welcome. We are super excited to kick off today’s session hosted by our incredible partners at Base 10. Before we get into it, I’d like to introduce myself.

I’m Geneva Scott, senior director of the Career Center. We’re so excited to have so many of you here joining us. In the chat, say where you’re coming in from.

We’d love to see some excitement in the chat. Just a couple reminders before we bring up our amazing presenters. At the end of the session, we will ask for your feedback.

We really appreciate you just taking a second or two to fill out the feedback survey. Really valuable feedback for us. Please utilize the chat to engage with each other.

Keep your kind of comments focused on the session today. Please utilize the Q&A tab to drop your questions for our incredible panelists. We will have a chance to get those questions answered.

And with those reminders out of the way, I’m excited to kick off our session with Base 10 today. And so facilitating our talk today is Sarah, who joined Base 10 in 2020, is the head of data visual- visualization. Could I say it?

And investment content. Prior to Base 10, Sarah was in content and product marketing at TripActions, a corporate travel management software company, where she created thought leadership, customer marketing, and product marketing content. She graduated Summa Laude from University of San Diego with a degree in marketing and a minor in leadership studies.

Aside from Base 10, when she’s not traveling or focusing on practicing her green thumb, Sarah, I need tips on keeping plants alive. She can be found walking around her neighborhood in the Marina District of San Francisco, with her Pomeranian dog Butters. I love that name.

Sarah, welcome and she is going to enjoy and introduce the rest of our panelists. Hi. Hi Geneva, thanks for having us and thank you all for joining us.

Like Geneva said, my name is Sarah and I’m our head of marketing at Base 10. We are a San Francisco based venture capital firm that invests in seed, series A and series B startups. You’ll hear a lot about that, so I won’t go too much into that, but I work with our investors to support their research efforts, which involves our investing strategy here at Base 10 and I also work with our research and data teams to understand how our technology stack can best support our investing team and our overall investing strategy.

That all gets packaged together and shared with key stakeholders and the success of Base 10 and the success of our community and that looks like founders, other venture capital investors and our limited partners who are the people that give us money so we can go on to invest in founders at those stages I mentioned earlier. We heard that many of you that are joining us today are excited to talk about venture and technology, so we’ve pulled together a panel to talk about both. I’m going to talk a lot about AI, which I know is a topic that’s on everyone’s mind nowadays.

So without further ado, let me bring on the rest of my team to join us today. So I’m very excited to welcome from our seed and series A team, Danny and Jacob. And we are also joined by Annie, who’s our software engineer who works on our data team alongside Robbie, one of our partners and head of data.

Thanks for joining, guys. Okay, now that we’re all here, why don’t we start with Danny? Danny, can you do me a favor and introduce yourself, how you got to Base 10, and a little bit about what your day-to-day looks like at Base 10 as an investor.

Yeah, thanks Sarah. And hi everybody, I am Danny. I’m an investor on the seed and series A team at Base 10.

I’ve been with Base 10 for a little over 2 years now. And prior to joining, I was at Insight Partners, which is a big venture fund based out of New York, investing in all things vertical software and AI. And what was the last question, Sarah?

My day-to-day? Yes, what your day-to-day looks like. Yeah, a combination of talking to a lot of great founders and, you know, hearing pitches, helping them reform pitches, as well as kind of leading research efforts and kind of thinking through what our next thesis in a given space is going to be and refining that.

And then a bunch of working with the founders that are in our portfolio to kind of help them strategize and and build great businesses. Awesome. Thanks, Danny.

I’m going to bump it to Danny’s team member, Jacob, who’s also on our seed and series A team. Jacob, same set of questions for you whenever you’re ready. Yeah, thanks for having me.

My name is Jacob. I’ve been, on the Base 10 team for slightly under a year and a half at this point. So joined, early last year.

Do pretty much the same exact thing as Danny, but prior to this worked at another venture firm called OpenView in Boston. I was there for 4 years and then another firm, called JMI in San Diego before that. And, I originally grew up in Los Angeles, so feels good to be back on the West Coast.

But yeah, in terms of my day-to-day, it’s pretty much the exact same thing as Danny. Awesome. And last but certainly not least, we are joined by Anirudh, who is our founding engineer, who sits on our product team alongside our head of data, Robbie.

Anirudh and I work quite a bit together, but I’ll let him give his own introduction. So hi, my name is Anirudh and I also go by Ani. I am originally from Delhi, grew up in India.

And, right now lead the, development of the Base 10 platform that a lot of our investors use to source diligence and kind of get intelligence on the latest investments that we’re making. Before this, I was at a YC startup as a founding engineer as well and I’ve seen how like a startup goes from like zero to one and then, you know, if there’s an exit or some sometimes it do, you know, go down and see the trials and tribulations of the startup world. And before that I was at Wolfram working on product with Wolfram Alpha, which I’m sure a lot of you guys have used.

So yeah, a lot of my day-to-day is just, you know, building this platform, getting feedback from investors, learning new ways that we can provide them intelligence and, you know, services so that the investors can go be the rockstars that they are. Awesome. Thanks, Ani.

We’ve seen in the the chat a lot of you have spent some time with Base 10 in the past. So welcome back. And for those of you who don’t know us super well, I’m going to explain a little bit about how Base 10 works and then we’ll go into a little bit more of a deep dive of investing strategy and about what we’re doing with AI both investing wise with Jacob and Danny and also on the product side with Ani.

This group is especially equipped to talk about AI since Danny and Jacob have had an AI focus on their research for almost about a whole calendar year now, I’d say. They’re still I have a meeting on my calendar that dropped this morning that we’re going to meet and talk about more stuff about AI on Friday. So I’m sure they’re just waiting to talk to me about that later this week.

But Base 10 Base 10 research is really important. We have two core pillars, research and purpose. I like to say that purpose is why we do what we do and research is how we do it.

So purpose is how we connect the success of our portfolio companies to the next generation of founders building startups. I’m sure that a lot of you on the call today are hoping to be future founders. But we work with a variety of organizations and nonprofits that make this happen including Coda.

So we’re very excited to that they decided to host us today and to be meeting you all. So big thank you to Coda there. But that’s the purpose side of Base 10.

The research side is how we find, identify, and define what makes a strong investment for the Base 10 portfolio. Every investor at Base 10, so Danny, Jacob, and the rest of our investing team spends time regularly with myself and our research team and Ani to identify potentially interesting trends, to meet companies and founders that are building in those trends, and to define what we call a differential thesis, which is why we believe a certain trend in software is worth investing in. And like I mentioned just a minute ago, Danny and Jacob have been spending a lot of time in AI this year.

We’ve had portfolio companies transition to being fully AI native, and we’ve also done some early investments in companies that were leading with an AI strategy from inception. So we kind of have a smorgasbord, I like to say, of AI companies at Bessemer. And we have a lot of different founders using AI to their advantage to build some really special companies.

But from a research perspective, to be effective, we spend a lot of time pattern matching and building frameworks to evaluate companies. So when we’re looking at AI, all these companies are very different, but they have a lot in common. So what’s making them successful, why are people buying them, and why should we be investing in these founders?

There’s also a lot of incumbent players and legacy players, I’m sure many of you have heard about and read about in the news, that are influencing how AI works. So that influences how our startups build and why we would or wouldn’t invest in them as well. So I’m going to start with Danny actually.

Danny, in our research meetings as a team, we spend a good amount of time discussing analogs and finding patterns in current innovation cycles. That just helps us compare to what we’ve seen previously in markets and with software companies. What kind of parallels have we drawn so far with AI, and what does this tell you about where we are in the maturity of the AI trend?

Yeah, 100%. I think without getting too in the weeds, probably the most high-level analog that we’ve talked about internally is the comparison of the AI boom to the cloud boom, right? And the cloud transition more broadly.

A lot of comparisons and similarities to be drawn there. Of course, it’s not exactly one-to-one, but the pattern matching can be helpful. I think one of the more interesting ways to look at it is if you break the cloud transition and the AI transition down into kind of layers of what the tech stack looked like for for development over time.

What you see is effectively three layers. You can break it up into how many as as you’d like, but I like to think about it in three. With the bottom layer being kind of the most fundamental of what do you actually need to be able to build all of this tech and infrastructure on top of, and the answer is the infrastructure layer, right?

So like semiconductors, chips, etc. And so for cloud you can think about this is all the semis companies. You can think about it as AMD, Intel, etc. But for AI you can think about it as Intel and all of the chip manufacturers there. What’s interesting is that in the beginning of both transitions, you see that most of the value accrual that happens, so like where investment is happening, where revenue generation is happening, starts at that infrastructure layer.

Which makes sense because you can’t do anything else without the infrastructure, right? The next layer you can kind of think of as either the step up where you start really seeing software or like the search engine layer in cloud. And so you can think of that as like all of the search engines that popped up, of course.

In the AI comparison you can think of it as like the model layer, for example. That’s kind of like the next level of value generation that we see. And then lastly is the application layer.

So for cloud you can think about that as all of the applications that end up touching the consumer and have gotten really big. So like Salesforce, Adobe, any of the businesses that you can think of as investors in software investing in in the last 10, 15, 20 years. For the application layer of AI, that’s all the businesses now that you’re starting to see like the Harveys of the world, for example, that are are, you know, the app the applied versions of some of these models that we’re seeing.

So what’s cool is that as I mentioned earlier, in the beginning of cloud you saw that all of the value is happening at the bottom, but by the end of the cloud transition all of the value really or the majority of the value is at the top at the application layer. And so we’re seeing really interesting parallels now with AI where most of the value is accruing at the bottom again with the Nvidias of the world and then the next level is the model layer and AI applications are still really early but we’re starting to see that transition happen where at the application layer is getting more and more exciting and invested in and seeing revenue generation and so we as software investors are getting really excited about the application layer of AI and that’s what we’ve been spending a ton of time on. Awesome.

Danny Jacob, I think this is a great conversation for both of you and you’ve already had a YC mentioned in the call so far because Annie did work for a YC company in the past. I’m sure everyone in this room is a little bit at least a little bit familiar with YC. I think demo day was actually yesterday for the last batch so it’s been very busy couple weeks with us because we’ve been meeting a lot of YC companies that’s actually a good amount of what Danny and Jacob spend their time on this time of year for the seed and series A company.

I Danny Jacob how has this batch of companies for the latest YC batch looked different to previous batches we’ve looked at? Have you seen any patterns when it comes to AI? How many of them are AI versus you know traditional SaaS or traditional software companies?

Yeah, I mean I I could take this for a second. I would say maybe like as of four batches ago, three batches ago, it was majority SaaS. And then you really started seeing like the initial AI companies come like a couple batches ago but as of last batch in this most current batch, it’s all voice AI.

I would say like that is the predominant theme by far. And you know for people that don’t know, I mean it’s it’s you know automating customer service calls or maybe it’s quoting automation calls for specific industries like logistics or hotels but that for sure is like the the biggest theme as of late. Danny if you want to add anything to that?

Yeah, I’d say voice and then And Jacob you and I have spoken about this a bunch, but like agents for everything, the agentification of X, Y, or Z. So it’s it’s definitely voice as a modality is has gotten really popular and there’s some interesting stats that you can look at that are somewhere online that kind of look at the number of voice AI companies per YC batch, but also just kind of the optimization layer and agents for different kind of back office tasks is something that we’re seeing a ton. I I think like relating a little bit back to the the prior question too about like where we are in the cycle.

You know, like you saw maybe once the all the foundational models came out in like 2000 like 2022, there was a lot of experimental buying in enterprises and people just seeing like, oh, this is a neat new technology, let’s set aside some money, figure out what’s possible. I would say over the last like year and a half, maybe two yeah, year and a half-ish, I would say people are getting much more stringent about making sure that the AI price they’re buying are actually one working and two showing real ROI. And as a result of that, I think a lot of YC back, you know, companies that Danny just mentioned, are really focused on like discrete workflows or true ROI or being able to actually say like, hey, we’re automating this function instead of just being like this, you know, experimental AI tool for an industry.

So, that’s been an like interesting evolution for the past, you know, like I said, 18 months-ish. Yeah, and to that point, that also means they’re focused on a lot of the markets where they know that buying is actually happening on a day-to-day basis and things aren’t just kind of getting stuck in proof of concept plans. So, things like legal, insurance, healthcare, we’re seeing a ton in those markets.

In terms of our portfolio, what have we been investing in that I would that you both would consider to be voice AI or we were excited about the companies because of their voice AI capabilities. Danny, I feel like you should take that one. Yeah, I mean, it’s kind of ties into to the point that Jacob just made, which is we’ve been really focused on voice AI businesses that have really clear and strong ROI.

And so, we are really excited about companies that basically from day one or around day one show really clear value proposition for the buyers that they’re going after. And so, as a result, we’ve made some what I think are really cool investments in healthcare and legal voice AI. More recently out of the latest YC batch in fleet operations tying in voice AI.

But point being it’s less about the vertical in particular and more about using voice as modality to show really clear ROI really quickly. Yeah. And and to get like a little bit, you know, finer there, it’s verticals where there are a lot of, you know, back and forth done using traditional phone calls.

I mean, that’s obviously a right for voice AI. And Danny mentioned healthcare, but like the I don’t want to steal your thunder too much, but like, you know, medical data retrieval. Now this is ton of manual workflows, back and forth, people calling each other.

I mean, that that is actually investment that Danny just made recently. Yeah, it’s it’s a company called Predoc that we just led the series A for. But exactly to Jacob’s point, that’s a really cool application where historically you couldn’t really do this like chart chasing workflow and like getting all of your medical record data in one place because you the process included a ton of faxing and a ton of physicians or healthcare assistants literally calling people in different offices being like, "Can you give me the chart for this person?" And so, voice is a really cool unlock for them because they can automate that whole process for the most part.

And basically as a result, they have a right to win the rest of the medical data retrieval process. So, I I think to Jacob’s point, it’s about like where can you plug in and take those manual workflows and as and turn them into an automated workflow and as a result, that gives you the right to win other workflows that kind of touch that workflow. I I would also add like that has resulted in a lot of verticalization of AI.

And like where we’ve been focused a lot lately is I mean we talked about AI for specific verticals, but you know, I think a lot of the differentiation and potential durability in the long term comes from these vertical solutions integrating with the rest of the softwares being used in that vertical and creating eventually full on, you know, like workflow automations within this industry and that’s sort of like what Danny was talking about in the beginning, which is agents, you know, now full on agents for discrete workflows. Awesome. And to steal more of Danny’s thunder, Prealize was really interesting company and we’re really excited about it due in part to the chief medical officer is actually a doctor and he’s a physician and he experienced exactly what Danny was describing firsthand.

And I think that that’s something, you know, we on the research side try to push our investors to find industries or applications of technology like this where medical stuff is not going to go away. Everyone’s going to have to go to the doctor. As there are more people, there are always be bigger use cases for that kind of data and that kind of retrieval.

So, that’s really sticky and that’s really important to be finding companies that are going to be able to have that staying power because they’re solving problems that are really long lasting. So, we’re very excited about Prealize. Danny knocked it out of the park with that one.

I do see a lot of you putting questions in the Q&A. We’re going to do Q&A at the end, so please hold on to those. We see them, we’re not ignoring you, but please if you have questions as we go on through the rest of the session, throw them in the Q&A.

We’ll make sure we get to them when we have time at the end. Danny and Jacob really quickly, many of the students that have joined us today and other students that we worked with in the past have expressed interest in becoming founders on their own one day. Since your team focuses on seeds at base 10, we’re often one of the first investors to join a cap table for a company for a software company.

So how are we currently thinking about companies that are building AI and evaluating them? What are exciting problems that we’re seeing being solved? What are we seeing when we’re excited about a seed stage investment versus when we decide not to go through with the investment.

Jacob, do you want to Do you want to take that? Yeah, sure. I mean, you know, a lot of the things that we look for are not necessarily different than SAS, but I would say there are some specific differences.

But I mean, in general, just like any other software solutions, it’s looking at like what does the current end market look like? You know, how large is it? Are What are the buy propensity like in this space?

Number two, what does the competition set and incumbents look like? And I do start like I do kind of think this is where it starts getting a little bit more important for the age of AI because I I think we sort of have this thesis that a lot of the AI solutions out there is not necessarily tech defensible yet. And this is really important where maybe if all the incumbents in a certain space are really fast and innovative, that’s a little bit scary cuz they can, you know, like build much maybe not faster than you, but they can build a solution and they have a much stronger distribution advantage over you.

And so that’s something to be really cognizant of versus there are a lot of end markets where maybe all the incumbents are all extremely old businesses and very sleepy or maybe they’re all private equity backed and you know, they they’ve effectively shipped a lot of costs out of R&D and so they’re not really innovating at all. Like that’s a perfect space to be building something new in. And then maybe the the last thing that we like Danny and I have literally just been talking a ton about lately is just a lot of these AI solutions start off with some wedge solution and we spend a lot of time thinking about where does this where does this wedge solution fit amongst the broader tech stack and does that wedge have the right to win additional workflows or not over time?

And I mean I think this is pretty critical because you know like some of the wedge solutions out there you’re kind of just stuck versus many of the other ones where maybe like there are very natural expansion points start attacking other workflows. That’s how you build a very large enduring business over time. Totally.

And to to add to that and to kind of just emphasize one of the first things that you said of like it’s not in many ways all that different from what we think about when we look at making a general software investment. I’d say the first principles are almost the same. It’s just maybe even more important now.

And by that I mean exactly what Jacob was saying. It’s like what is the actual vision for the business is maybe a different way of putting the wedge comment that you made because I think in a world where it’s really becoming much easier to grow pretty quickly from a small business to a not so small business with AI functionality because the demand is so high for some of these things. Where do you go from that kind of initial landing product and what is the thought process that you have behind like why does the market need me versus others?

And that doesn’t need to mean like your tech is differentiated necessarily, but if it were me I would be thinking through like okay, well, this is a really cool wedge product that I can get to let’s say like a million in revenue with. Awesome. From there though is that going to be able to sustain a really large outcome for me, a venture backable outcome, or do I want to try to move into different parts of my customer’s workflow?

And if I do, is the starting point that I began with a good jumping-off point for that, and why? Like, why have I really gained that right to win with that customer? And some of the businesses that we’re getting really excited about are A, being very thoughtful about what that starting point is, but B, part of the reason that the starting point matters so much is that if you can unlock a new customer segment in a market, for example, that hasn’t necessarily gotten as much love from the software ecosystem because maybe they’re down market or a little smaller and not historically like really feasible to serve just from a unit economics perspective, but using AI you’ve made that feasible, you might now have real right to win with that customer base, and you can kind of keep going and building software for them.

So, that’s one example of how I think about it, but I think getting creative and also having real vision towards who you’re building for and what they really need and why you feel like you have a unique right to win is is probably key to know. And maybe just to give like a tangible example of what what we mean by you have the right to win. Let’s say you decide to build a, you know, a voice AI solution for picking up customer phone calls for let’s say like restaurants.

But when you start, you know, maybe interviewing restaurant owners, they mostly complain about, let’s say, accounting being like the biggest pain point. If you’ve built that voice AI AI solution for answering phone calls, you don’t necessarily it’s not a natural expansion point to start building accounting solution. Like, the owner would never really kind of correlate the two.

And so, like, that’s, you know, you really don’t have the right to win that accounting workflow or job to be done. So, you know, you got to kind of start really thinking about like what are the people in this market saying, and where does my initial product make the most sense to expand into their biggest issues. Totally.

Awesome. I think we’ve been talking a lot about how we evaluate software companies and how we think about technology from an investment perspective. But Danny and Jacob, I’m sure you both agree that we are pretty data-heavy investors and we lean a lot into using technology internally, to help inform our investing process and make your lives easier.

We will have you look at more companies than you would be able to otherwise. And this is also why Ani is joining us today. So I want to talk a little bit about, you know, from an investor perspective at Base 10, how do you both work with Ani and with our product suite?

How do the tools that he helps design, like help fill the investment team, with more insights, more data, and just an ability to kind of create this flywheel for our investing strategy? Danny, you want to go for it? Yeah, sure.

I mean, it’s it’s almost hard to summarize because there are so many ways that it touches my work on a day-to-day basis, but like, I feel, with some of the tools that Ani and the team have developed, it’s just like being an investor, but with like some level of superpowers. And by that, I mean like, I can click into a company URL and know like how much money they’ve raised, like what, you know, thesis they might fit into for us, like who the founders are, like it gives you real data on like how fast they might be growing, etc. And similarly, that’s just on like a company-by-company basis, but Ani’s developed tools that also kind of like systematically help us identify based on our preferences and themes that we think are really interesting, here are the companies that you should be talking to and here’s how they’ve been generally doing and here are some of the industry trends that they might be riding off of, etc. And so, it’s effectively taken a lot of what I would think of as like the manual work or things that we might not even really be able to get insight to without, speaking to the business and then trying to aggregate all of that data ourselves and kind of giving us the the the information that we need to be able to make an informed decision about, you know, who we should be talking to and where we should be spending our time and it’s kind of your fingertips. Audie, I think that’s probably a good time to pull you in.

I know we already did a little bit of an introduction for you. But can you tell us a little bit more about why you decided to come to Base 10? It’s a really I think different or unique engineering type role given the problems that we’re solving here.

But I’d love for you to share with with the session a little bit about the opportunities offered this team and why we’re able to benefit from technology and from tools that you’re building. So yeah, I I mean I’ve been in San Francisco for like about eight years now and then before joining Base 10, as I said, I was at a YC company and I literally met the founders of the YC company at a WeWork where we would just be like hanging out, chatting, talking about tech, talking about new stuff. And I literally like joined them because I used to play ping pong and chat tech with them the whole time.

And and I just love, you know, startups the the the whole process of people, you know, building these huge businesses out of like small ideas and the work and the kind of insights they have to come up with. So I’ve always enjoyed the startup space. And so after as I was working at the startup and things things were not really working out at the startup and we were all trying to explore alternate opportunities.

I was able to meet Robbie who’s like one of the partners at Base 10 who has this cool vision of like, why don’t we build a, you know, like a complete product tooling suite that is focused around giving, as Danny said, superpowers to investors. And as someone who just enjoys startups and and and the pursuit of finding like cool solutions to like complex problems. It felt to be like I was so excited to build something that looks for those.

And this was like kind of a meta level of like building product, but it’s such a fun fun problem and something that I don’t think I would have had the opportunity anywhere else. And let alone you know, a venture from betting on putting tooling first to like empower investors. So it was a it was a opportunity that to be honest I could not I could not, you know, let go.

And also talking to Robbie, there was like an instant synergy where like he has these cool vision he has this cool vision that I really synergized with and I had so many ideas. And like our our first call where we were supposed to like just get to know each other and he was supposed to introduce me to the role was scheduled for 15 minutes. And then it ended up being like about an hour and a half and he was like, "Oh I have another meeting to go to, but let’s continue the chat later." So that that that goes on to show how much like there was an instant instant like synergy.

Yeah, absolutely. I remember Robbie being late to that other call. He’s like, "Sorry, I was talking to somebody I want to hire." So we let it go that one time, for sure.

On you a lot of the students that are part of the CodePath program that we’ve met with CodePath over the years either are majoring in an engineering type role or have a background in computer science. How would you describe this role and what you’ve been doing with the ton is different from a traditional engineering role at a software company or from what you’ve done previously? And also like where is it the same?

Where are you kind of using the same skill set and the same problem-solving skills? So a lot of the work from an engineering side of things that I have to do are very full stack work where I have to do things from the ground up and build completely new tools. I have to build a front end and a back end and go full stack and do everything from scratch because, you know, this is a completely new product suite that we’re building.

But, a lot of the things that help me do it better is a lot of product work. Like kind of talking to the investors, understanding what their day-to-day looks like, looking at different workflows. Our research team is one of the biggest kind of drivers of how we’re like looking at companies.

And so, going deep into that research, being there every like research dive, understanding what it is about, and then kind of translating what was talked about in those meetings, and the insights that our investors have driven into like things that we can add into our product that can like allow them to find those things much faster. And so, to go on a little bit of a tangent, the way we do product at Base 10 is we have this one thing called the extension, which is what Danny was talking about where she just clicks a URL, and she she can like see the entire information about that company, how much they’ve raised, how many employees they have over time, who’s funded it, who are who’s on the cap table, and all those things. And that is like our, you know, point, like if you were to draw an analogy, a point of sale intelligence.

Like that’s the touch point that the investor has with the company, and getting every context for that company the investor needs so that they can like kind of picture it in regards with other companies that they’re thinking with with that kind of investment. And the other thing is a kind of intelligence plus sourcing tool where we take insights that investors have come at through their extensive work and, you know, insights and thinking through different trends, and they give us like these pieces. And they allow them to apply the that thesis to like an extremely large set of companies using like LLMs to see like, "Hey, let’s scrape all possible websites for a company and then we will kind of match it against this thesis.

And once we find a high conviction thesis, we can like kind of surface it to the investors for them to do their things. So, it allows an investor to have a much wider field of view of like how they’re going about finding companies or sourcing companies than would have been possible before. So, it in these things like finding these two like foundational things that we have to work on.

And now we have we’re going to add another one, which is a, you know, kind of an AI system that allows them to like chat through with how does how like select a bunch of companies and decide how does this translate to this thesis and which is the strongest kind of one that attaches to the thesis and things like that. So, finding those fundamental like blocks that we have to build and separate them out and have some sort of utility to it required a lot of product work where we’re just like just in the weeds with the investors, not doing any computer science work, to be honest, but more just like research and being curious about single thing it’s like proper things. But at the same time, like, you know, the infrastructure that goes behind doing 100,000 scrapes on so many websites and storing those scrapes, creating embeddings, and then finding an exact thesis match is like a hard core computer science problem where like, you know, within the resources that I have I have to find those fits that like a really good accuracy and not spending a lot of resources like I can’t just brute force search everything.

So, there when it comes to like the smaller problems that we arrive at after doing the product work, it comes back to the raw computer science skills of like how do we like problem solve that particular technical problem. And just a funny example here of Ani being like the you know, a real product owner. He had like a little beta group and he asked us basically to shadow us on how we do our job and evaluate a company.

And we were, you know, usual way is we’re on some company’s website then we try to figure out how much money they raised so we go on a PitchBook or Crunchbase and then we try to find the company on LinkedIn to find the CEO. And then you would go on Apollo or something to get their email address and he was like, "This is ridiculous. This doesn’t need to be four different websites.

We can do this in one." So, it’s yeah, it’s thinking through things like that. Okay, awesome. Cool.

I So, we are about 25 minutes from the end of this panel. Do you want to call out really quickly that Jacob does have a call with the founder in 10 minutes so I told him he could go a little bit early just cuz we want to make sure we don’t leave that founder hanging. But we will get through a Q&A, I promise.

I did want to make sure that we had some time as a group to talk a little bit tactically for everyone who’s joined today about things with your career and about engineering specifically and about just a general career advice before we get on to the Q&A. And like I said, we have a a bunch of questions coming in. Please keep them coming.

We’ll try to get to all of them. But Jacob, really quickly, let’s start with you. What’s one mistake you see early-stage founders making again and again that you would advise someone, like someone in this group who’s joined us today, who’s going to be starting their own business, what that you would advise them to avoid?

Getting overly attached on the first idea and even with, you know, evidence or data that, you know, again, like this wedge is not necessarily the best wedge or, you know, it’s it’s not the best way to go from a single product to multi-product company. It’s falling into that trap, I think, is is a really big issue. And then, you know, like a a lot of the founders that we talk to, what we are really discussing is not like the product today, but it’s like the road map over time.

What is the greater vision for this thing? And being able to articulate that clearly I think it’s really really important. So, have have that broader vision as you’re thinking about what you want to be building.

If I can add to that really quickly and I think this is a good like flush out point cuz I completely agree with Jacob, but the other aspect too is trying to stay focused is really important. And what I mean by that is there are so so many really cool things that you can go and build. And even if your first idea isn’t necessarily long-term the one that you’re going to end up sticking with.

I think it’s really easy to get distracted in the early stages such that like you never really get to an answer on what you should be pursuing in any direction. And so, I think picking a direction, getting the data points on whether or not that’s a good route to go, and then from there moving on. And you obviously while you’re getting those initial data points can be taking note of other cool ideas and opportunities that might come up.

But I think trying to get that initial data points on whether that first idea is a good one and then if the second idea is a good one or not. Because otherwise I think if you try to bite off more than you can chew, it can get a little bit unruly. Danny and Jacob, I think that that’s also kind of leads into a good question about what you’re thinking about when you’re starting a software company in particular.

How do you how have the both of you kind of seen technical founders best complement themselves with non-technical founders? We I think we’ve kind of invested across the board. So, we invest in very technical folks.

We’ve invested in teams that are entirely technical and some where the founders aren’t technical at all. So, how have you seen those dynamics kind of play out between founders? And again, how you would advise someone who I think a lot of the people on this call have technical backgrounds and would be considered technical founders one day.

How you advise them to kind of go into the world with that with that mindset? Yeah, I think the keyword is complement, right? And being really aware of where you feel that your strengths are and where your blind spots might be and going and finding a co-founder that you think really highlights those areas that are important that need to be filled.

And again, like there’s really nothing wrong, I think, with knowing what your strengths are and wanting to lean into those strengths. I think sometimes founders feel that they need to be everything all at once for everybody. And in order to be the perfect founder, you need to be great at sales.

You also need to be great technically, and you need to be great at pitching investors. And the reality is, not everyone’s great at all of those things. Some people have superpowers and and spike in different areas.

And so knowing where you spike and being able to kind of fill that out with your co-founder is, I think the best thing that you can do. Okay, awesome. Anny, let’s go ahead and let me ask you one more question before we go to the Q&A.

I’m still reading through all of your great Q&A questions, everyone, so apologies for the distraction. Anny, what’s one piece of advice you would give a software engineering student who hasn’t graduated yet, but hopefully will be soon, who wants to build an AI? I mean, this is almost things, but one of the most important things that I would say is be really good at using AI tools.

A lot of engineering has been completely transformed as to like not just cursor, but and cloud code and just in general using AI really well to kind of be faster than like the expectations of engineers have like gotten higher, way higher, than 2 years ago. And it is because of how all these AI tools are augmenting engineers. So, one of the things that a lot of people kind of stress on is how good you are at using those tools and how good you are at debugging cuz they’re not really good, but there’s like specific frameworks you can use like such as the B mad method with Claude code to like get to the end state that you want to get at as fast as possible with the least amount of debugging time required.

And so that that I think I’m I’ve seen how even in a lot of interviews and a lot of hiring decisions like just the proficiency in using AI in your coding workflow has become one of the most important factors. And it it a lot of companies are also just mandating like you have to use AI like in your coding workflow even if you don’t think it’s good because they see that there’s quite a difference in terms of output when you are not using AI at all versus when you are. So I I feel like it’s not just now a just a thing that you use to like get faster, but it’s just a necessity and you got to hone in that skill of like how good are are you at using these systems that that you’re comfortable with and how much do you know how to make them work in your particular context really well cuz it is different it is different for different kind of tools and like the Claude code like how good are you using plan mode or like debugging?

Are you letting the take full control or are you writing a good rules file? Like those are important things I feel like that someone looking for engineering jobs today needs to be just really good at. Awesome.

Thank you Annie for that. I feel like you Well, one of the reasons I brought you here is cuz you have that perspective that I feel like the three of us can’t because we don’t have those hands-on skills. As much as Danny and Jacob talk to founders every day, it’s just it’s not the same.

So thank you for being here and thank you for that. I do think we want to get started with Q&A. So we do have quite a few in here, but I wanted to start with one of the top voted ones.

And CodePath folks, please let me know if I’m doing anything wrong here. Yeah, here we go. Perfect.

How do you differentiate between hype and truly transformative use cases when it comes to evaluating AI opportunities? Danny, Jacob, I feel like this would be good for the two of you to offer an investor perspective, but only if you have anything you want to add, please jump in. Jacob, do you want to You have to jump in.

Yeah. Yeah. This is such a good question cuz it’s literally what we think about all the time.

It is. But, I I would say like it really has sort of come down to what what we alluded to in the very beginning, which is you really have to think about what is a value proposition and ROI of the solution. Like we mentioned, like a lot of experimental buying has gone away from, you know, larger enterprises.

And so, these companies now have to get much better about articulating exactly what they’re providing, which is either, you know, cost savings, time savings, or revenue generation. And I think when we think about something that’s truly transformative, you know, it it’s companies that are very very clearly able to tie their product to exactly how a you know, a business will be changed. I.e., you will you know, generate X% more revenue compared to your old ways based on this.

A lot of the hypier, you know, companies are usually a little bit vaguer about this, but, you know, I think there has been a lot of, you know, supposed value creation so far in some of the hypier, earlier AI companies that you know, maybe come out in 2021, 2022 that may in the future start seeing some cracks because a lot of their customers just aren’t seeing true value yet. I won’t name names or or mention any specific ones, but you know, I think that’s something that we’re pretty keenly aware of lately. Danny?

Yeah, just to add to that and maybe put it in a little bit in a different light because I think talking about ROI can be a little bit like how do you know what ROI is 100% a lot of the time? And I think a different way to look at it is especially as you all think about starting businesses, what are your customers saying? And we as a part of our diligence process get on the phone with customers of these companies, right?

And we ask them about the product and we hear about how they’re viewing it and how they use it and, you know, the value that it’s adding to them and that’s kind of how we get at ROI a little bit, too. But one of the things that I love to hear and like this is the holy grail, like if you can get to this from your customers consistently, you’re in a really good spot is I don’t remember or I don’t want to remember what my workflow is like before using this tool. This changed how I work or this has given me back hours or it’s saved me X dollars, which is again kind of getting at ROI.

But hearing real customer love and appreciation for the product is something that I actually really don’t think anyone should take too lightly even though it sounds a little bit more touchy-feely. But customers and their willingness to vouch for the fact that they don’t want to rip you out of their tech stack because it would make their lives worse is I think a really interesting way to differentiate between hype and transformation. Because yeah, once once your workflow’s been transformed, you don’t usually want to go back on that.

Versus a hypier solution may have a lot of promises, but not necessarily be able to deliver on that real that real workflow change for the customer. Awesome. I think we want to pull in a question that we got from CJ.

There we go. How should one stay up to date with the emerging trends we see and how do you discern between smoke and actual fire? I do want to pop in and answer this one really quickly just because I think for any of you that do go on to pursue careers in venture capital, it’s super important that you stay up to date with trends in technology whether or not you take on a role like Annie’s where you’re building product and working more on the platform side which is the side that we describe as like what you do to support investors.

So that’s what I do. That’s what Robbie our head of data does. That’s what a ton of other folks on our team do any kind of a supporting fashion to the investment team.

Staying up to date with trends is generally one of the better things you can do for interview preparation. Danny, Jacob, Annie, I don’t know if you guys have specific stories but I almost every interview I’ve ever had in my career, someone has asked me something along the lines of what’s a really exciting company that you’ve seen recently. Since my my background is in marketing, I got asked in an interview for Base 10 like what are some of your favorite startups that do branding really well or have really great websites.

So being able to kind of think of that off the top of the head really comes with repetition of reading TechCrunch, subscribing to newsletters that investors read, having like alerts on your phone for for news sites about technology, and really making sure that you you keep up with those headlines, keep up with who’s been fundraising, who’s been raising money from VCs, who’s acquiring who, who is leaving Open AI to go work at Meta and all that good stuff. Like keeping up with that stuff is is really important. But yeah, Danny, Jacob, Annie, what advice do you have for students here?

Yeah. I agree Jacob. I agree Danny.

Yeah, I was just going to say like I I I think the newsletters, publications, all of that is super important and I completely agree. I think the other thing that I’ll add is just like talking to the people around you. What are you using?

What are you seeing on a day-to-day basis just getting you excited? And they don’t have to work in tech, honestly, like a lot of the questions that I ask my friend friends are to the ones that don’t work in tech cuz I’m like, well, what piece of software have you used recently that’s kind of cool or that’s changed the way that you’re working. And more often than not there will be an answer and that’s a thread to pull, right?

And so, being able to do that pattern matching both through what you’re seeing in your day-to-day life, but also in, you know, again, the newsletters and other things that that you’re looking at, you know, the type of startup that’s getting funded a lot of the time has a lot of similarities to the last batch that you just saw. And so, being able to kind of look at what the similarities are, but also discern what the differences are kind of gives you a thought into, okay, well, what could be changing in the industry and why. Yeah.

Maybe to rehash this again, but I’m only doing it cuz I think it’s actually really important. I probably spend like yeah, maybe like the first hour every day just just reading through newsletters and news websites. And so, one, subscribe to the Wall Street Journal and New York Times.

And then there are a bunch of like VC specific ones like, Axios Pro Rata, Fortune Term Sheet, and a few others. And just like literally dedicate time to to read through that. And then, actually I forgot the other point I was going to make, so I’ll just stop there.

One, one of the, I mean, the podcast, the newsletters, they’re of course kind of the cornerstone of like how we’re mostly beginning all these updates, but what there are so many of them. And so, one of the one of the tricks I found to like be able to consume more of them is to just leverage AI to do it better. So, this is just a personal hack, but I I use this tool from Google called Google Collab, where I will keep like, sending some of the interesting newsletters or podcasts or YouTube videos as well.

And YouTube is some great creators on YouTube who are like really good at keeping you especially for engineering. There’s a lot of creators that are always constantly pumping out really good content to keep you up with like what’s happening. And so I will find all of them put it into one collab notebook.

Sorry, not a notebook alum, not collab. A notebook alum. I’ll put them into one notebook alum notebook and then do an audio overview and it basically creates this long podcast which combines all my all the important newsletters and all the important things and emails that I’ve gotten one day and just gives that to me in the morning as an update.

And that honestly has just been like pretty transformational in how I keep myself up to date with like all the things happening. But yeah, leverage I I would I would highly suggest leveraging AI to be really good at continually consuming that. I actually just remembered the other point I was going to make.

It it’s if you are trying to get a job in VC and I’m sure this will actually be applicable you know, if you’re trying to apply to start up or something. But as you’re reading through all the newsletters, keep a little like running log of companies that you think are most interesting and just update that as you know, maybe something becomes less interesting over time and you find something newer, but keep a little log of that and just write little notes of like why you think it’s interesting and you know, maybe what they’re doing better than a competitor or something like that. And that’s always good like conversation fodder.

Totally. And if there’s a backup like industry trend as to why you think that company is interesting, that’s like bonus points and also super important. Cuz it’s one thing to say this company is cool because it raised money from an investor that I really like versus I think this company is really cool because it’s riding off of the tailwinds of this change that’s going to impact this market in this way.

That’s an answer that’s really insightful. This may be a like a maybe a little bit of a layer too deep, but a lot like Danny just said, when bigger investors or VC firms invest in some of these companies, they’ll also write blog posts or articles about why they’ve invested in these companies. They’re not super fun to read, but if you stick them in notebook LM, you can probably save yourself a little bit of time like Annie said and just try to understand you know, why some of these bigger investors are investing in these companies.

They’re not always right. Some of them are hits, some of them are misses. But being able to also kind of like learn the language of how folks talk about software companies, well, even if you do not get a career in finance, if you go on to start your own startup, you’re going to have to speak the language of VCs and of investors if you want to pitch your company.

Danny, Jacob, Annie are, you know, hearing pitches all the time. So being able to really understand why you’re excited about the product that you’re building, why there’s a huge market tailwind that’s going to push your product forward, and why someone should give you millions of dollars to make it happen, that’s a really invaluable skill. Even if you are, you know, not pitching a startup, but you’re pitching yourself and why you should join their company, having that skill is super important, but reading and consuming that content is just one way of getting there.

I think this is a I thought this question was super interesting because I think Annie, it would be great for you to kind of like talk a little bit of how Base 11 gets its data and its resources. And then Danny and Jacob, how Base 11 is is better than these data sources. So is Base 11 like a PitchBook or a Crunchbase?

Base 11, I think sits on top of both PitchBook and Crunchbase is like of course a lot of the data that we have we get is from a lot of data providers like PitchBook, Crunchbase, and there’s many others like Prequel. And so the thing is these are all tailored to just give you everything at once. And the the problem is that we have a very specialized thesis and specialized way of functioning which is very research focused that is very unique to our firm.

And so, Base 11’s job is to take all this data and augment it in a way that makes sense for our ways of functioning. So, we have a notion of like business like of megatrends, where like we will define what is a megatrend and have like a good semantic understanding of what that means. And then, what Base 11 does is take that semantic definition of a trend and then grab data from PitchBook, grab data from Crunchbase, grab data from all those sources, and kind of contextualize the best possible companies within the framework of that trend.

And it’s not just, you know, there’s a lot of scoring algorithms that we use, there’s a lot of matching algorithms we use, we use a lot of AI for semantic searching. So, Base 11 does it sits on top of them to kind of provide more intelligence from data that is just that we receive from these sources. I hope that answered your question.

And then, I think we have just a couple more minutes here, so I think I want to hear, I think from all three of you because I think this is a really important question and one that we talk about almost every day. Do we think that we are in an AI bubble? Not the bubble word, I hate that word.

Or is it something that will continue to push the industry in the long term? It’s a question we hear almost every day from our investors that give us money. So, Danny, Jacob, Annie, Yeah.

Bubble or no bubble? I I actually think this ties in really nicely to what we were talking about at the very beginning of the panel, right? Which is how do we think about analogs for the AI transformation.

And when I think about it again, I compare it to the cloud transition, arguably of larger scale, I would argue, my view. But one thing that you saw in the cloud transition were of course like hiccups along the way. It was a long and meaningful transition, but this this shift from manual workflows V1 software workflows into AI is generational.

My view is that we’re still very much in early innings. And again, like if we’re thinking about this on the scale of what the transition to cloud was, again, we’re still in incredibly early innings. And the bubble word comes from like are we running up value too soon in the cycle?

And I actually think that the more relevant question is like where are we in the game? Are we late stage or are we early stage? Is the transition almost over?

Are we just at the very tip of the spear beginning? And the data again, from kind of like where we’ve seen value accrue and you know, where we’ve seen, you know, revenue start to trickle in would imply that we’re still really early. And again, that won’t mean that there aren’t hiccups the same way that there were in the cloud transition, but you still saw massive winners from the cloud transition that I think will reach even larger scale in the AI transformation.

That’s my take. My take is if I had to answer it quickly is is not a bubble unfortunately because it also definitely feels like it at times. And when when like if I were asked, I think even like 6 months ago like what inning are we in in like baseball nine innings, like I would probably said like second inning or so, but I think the more that we spend time like talking to customers and some of these AI companies, it’s maybe actually more like 1.5, like even earlier than I actually thought in terms of like a lot of things are actually really working.

And a lot of people are actually getting a lot of value out of it. And turns out there’s actually still so much more that these companies could be doing and so, you know, like in in a kind of a scary way actually do think yeah, maybe we’re still pretty early on and all this. So yeah, I would say not a bubble.

Hopefully that doesn’t come to bite me in the butt, you know, couple years from now, but yeah. Awesome. I think that’s all the time we have for questions.

Geneva, I’m going to hand it back to you. Awesome. Could we please give a massive thank you to our incredible panelists in the chat for their time today.

I learned a lot. This is actually the second time I’ve heard you all talk about this and I feel like I learned more today and really appreciate it. And so just a couple reminders as we close out.

Again, show your gratitude for the Base 10 team in the chat. You please go back to the schedule to see next sessions. I believe next up we have AWS and Google.

You’ll see at the top. The student survey, we really appreciate your feedback on the session. And to close us out for our session today, we’re going to go to a brief video from our host Bobby D.

Thanks everybody for attending. See you in a session soon. Now wasn’t that a great session?

Oh my gosh, these breakout sessions are fire and I want to encourage you to attend as many as you can. But let me give you some information before we continue. So please take a moment to complete a brief survey to share your feedback.

Remember surveys are what’s going to make us get better, but also let us know what we’re doing really well. To see what’s coming up next, you can head back to the main stage and check out the schedule. See you soon in the next session.