How Delphi Raised $16M from Sequoia to Build Digital Minds

In 2014, I was gifted a book called How to Create a Mind by Ray Kurzweil, where he talks about how one day you'd be able to recreate someone's mind using computer science. A couple of things we've learned along the way in terms of the product— initially we started consumer first, and then we kind of realized, okay, in order to really make this a mass-scale platform, first we need to focus on the people. Something fails not because the idea is bad, but because it hasn't been cooked enough. You haven't iterated enough. And I was like, Always had that concept in the back of my head. Like, okay, we're false failing right now. Like there's no way this idea is a bad idea. We're just false failing.

Welcome to Just Go Grind. I'm your host, Justin Gordon.

Today's episode is brought to you by Relay Human Cloud. When you need great talent fast, Relay connects you with a curated pool of global professionals who are fully vetted, ready to interview, and can start within days. They handle payroll, benefits, compliance, and secure office space so you can focus on building while saving up to 75% on labor costs. No recruiting fees, no long contracts, just top talent on demand.

Ready to scale smarter?

Go to relayhumancloud.com, book a quick intro call, and use code LAGRIND for $500 off your first invoice. Let Relay power your growth so you can just go grind. Again, that's relayhumancloud.com.

Let's get to the episode. Dara, welcome to the show.

Justin, thanks for having me.

I love what you're doing with Delphi. When you reached out, I finally got back to you and was like, this is going to be awesome. I had to chat with him. So for people who don't have context, tell me a little bit more about how you describe Delphi today.

So Delphi allows you to create a digital version of your mind to scale your expertise in an interactive way. If you think about like a blog post or a book or a YouTube video or a podcast, these are ways to scale your knowledge in a passive way. People can consume your content passively. Now you can create something that people can consume your content through conversation, interactive and personalized.

It's so interesting. And I was actually playing with it right before this. I was talking to Genghis Khan, going back to like an old school, because I'm like, he built an empire. How to build an empire is what I want to do. And it's very interesting. Like, the number one thing was like, unite your people. It's like his first response. I was like, this is fun to play with. I can see the value property from many different ways. But I always want to know, like, how did this all start in the first place?

I mean, the dots always connect looking backwards, so I'm going to tell you the story, but it wasn't as obvious at the time. In 2014, I was gifted a book called How to Create a Mind by Ray Kurzweil, where he talks about how one day you'd be able to recreate someone's mind using computer science. And that got me really interested in AI. I was an engineer for a while, and then I started my first company in 2021 when I got early access to the GPT developer list. And I was like, this is going to change the world. I should just like do things in this space.

Yeah.

And I started my first company. It was an AI assistant for shopping. It was like buy anything with a text. And I was reading a book about my grandfather and I really wanted to get his advice because he was this really, really successful entrepreneur in Iran before the revolution. And obviously a book can't give you advice. And he had had a stroke a couple years prior. But I had this realization that LLMs kind of matched the description of the technology that Ray Kurzweil talked about in that book. And I ended up creating a digital version of him that I could talk to. And it was surreal. And I posted it on Twitter and everyone was like, you're crazy. This is immoral. You shouldn't do this. And I was like, okay, maybe I'll put that on the side. And that first startup got acquired, but that experience just got me really thinking about, damn, this would be so useful. I've never really had a mentor in my life and it would be so useful to be able to learn from any of the best people in the world, but they're constrained by time and there's no way to scale yourself. And so I had been working at Open Store for Keith Rabois, and I met my co-founder there, and we just kind of got really obsessed with the idea. And we started the company about like a month before the launch of ChatGPT.

Crazy timing on all of that. One of the things with your co-founder, since you mentioned it, is what were you even talking about? Like, how did you guys become like friends and want to work on something together? Because I mean, it's so hard to find the right co-founder, and you bootstrapped as a solo founder with your first company. What was it about your co-founder that you want to work together?

Yeah. So actually, when I sold my first company, I was like, I want to start another company, but this time I'm going to do 2 things differently. I'm going to work on something that I'm absolutely obsessed with, which the Delphi idea I was getting obsessed with. And 2, I want a co-founder. Where can I have highest probability of finding a co-founder that I'm going to get along with? Okay. The kind of person that's going to move to Miami to work for Keith Rabois, who's kind of a controversial figure, is probably someone that I'm going to get along with. So I moved to Miami for the purpose. of finding a co-founder. And at first, Sam and I were kind of rivals because we were both overachievers, trying to be the best guys in the company, working really hard. And then we started a book club together, and we just— he was really curious. One weekend, we decided to just go to Santo Domingo together, and that is really not a place for tourists. And I think Sam Altman said somewhere that the best way to figure out if you should be a co-founder is to go on a high-stakes trip with them. And I think that kind of solidified it for me, like, okay, this guy's great. He kind of complements my skill sets, but we are grounded on the same value set. It's kind of like getting married in a way.

We're going to get into a lot with DeFi and your story and journey with that. Taking a giant step back, why entrepreneurship in the first place at all? Has it always been the case with you? What are you drawn to with it? Like, I want to know more about your background in that way.

I mean, I think there are a lot of reasons why. I mean, the most obvious is the family. You know, my grandfather had 30,000 employees in Iran at his peak and You know, then he was kicked out of the country, he lost everything. So I grew up always hearing stories about like the amazing life my family had in Iran. And I was like, wait, I wanna, I wanna do that for you guys. Like, why, why don't we have that? I mean, we have a good life now. My dad's an entrepreneur. So I think at an early age, also, I grew up in Texas, like Middle Eastern kid in a very white school, experienced a ton of racism. People would call me a terrorist. So there's probably a bit of a chip on my shoulder. But yeah, I think it just, it's fun and I also like what I'm doing.

Yeah, no, that's interesting because that part of it too, I have many different friends who are founders and it's like, why are you building this thing and choosing this insane path and like trying to be super ambitious with something is fascinating to me and what they choose to work on. Because if you're building your life's work, I mean, you're going to spend so much time. Like I was reading about your culture and everything with Delphi and like, yeah, you're working 6, 7 days a week, but people want to because you guys are you want to build what you're building, which is, which is interesting. One of the things I'm curious about too, in the early days, like what did you think Delphi was going to be exactly? You had some experience with AI and everything. Like what do you think it was gonna be as a starting point?

It's funny, like our pre-seed deck is pretty much what we are right now. So like we, we've pretty much stuck the path. I think a couple of things we've learned along the way in terms of the product. Initially we started consumer first and then we kind of realized, okay, in order to really make this a mass scale platform, first we need to focus on the people. What would incentivize someone to actually create a digital version of themselves? We want to be the verified platform. We don't want to be Character AI. So then we kind of shifted our focus towards the creators and the authors and the value that they might get from scaling themselves. And it didn't work for like a year. Everyone was telling us to pivot, and it was pretty dark. But I was very confident in the product because I was using it. And also, there's a book called Loonshots, and in that book, they talk about this idea of a false fail. Where something fails not because the idea is bad, but because it hasn't been cooked enough. Like, you haven't like sat on it and you haven't iterated enough. And I always like always had that concept in the back of my head, like, okay, we're false failing right now. Like, there's no way this idea is a bad idea. We're just false failing.

What is it about that time period, like that year, that you knew it wasn't working? Like, I mean, certain numbers you weren't hitting, metrics you weren't hitting. What did you care about? Like, what were you trying to see in that first year that you'd be like, oh, this is great? That you weren't quite seeing yet?

Any kind of traction or like we weren't getting really signups or the people that we wanted to have on the platform, like some of these famous people, just like were not interested. And I think a big learning that we had is the kind of company that we're building needs to be a brand. Like it's not just a software company. If people are gonna trust you with their identity and their data, we need to be like a brand that people wanna associate themselves with. something that people aspire to. You're immortalizing yourself. We're honoring your mind so that other people can reach it, and that it's gonna last 500 years from now. You're trusting us with your digital legacy in a way.

How'd you get to that point then, in terms of that? What was it like that you flipped the switch with it?

A lot of iteration. The product got good enough. We made some rules. You can only create a digital version of yourself. You have to submit a photo of you holding your ID. We built guardrails to protect against hallucinations, which a lot of people were afraid of. We hired a designer and made the design really nice and meaningful. And then finally, our first big customer, his name is Brendan Burchard, he's a coach, and he reached out to me on Twitter and he's like, dude, I've been thinking about this for so long. Google tried to get me to do it and I didn't want to give up my data to them. And then I was paying some developers to put it together, but it wasn't really useful. And you guys are doing it and I own my data, let's do it, let's go. And as soon as he got on board, it was almost like a chain reaction of social proof where one person led to the other.

Yeah, you just asked for those, you kind of showcase him and then you can obviously ask him, like, we already have on the platform him. And a lot of people will not know him in this sphere. I know of him from that sphere as well, of coaching and stuff. He's massive. He's massive. There's a massive audience.

He's been so critical to our journey. He's great.

So where did it go from there in terms of getting these more of the B2B side and getting people with bigger audiences on the platform?

Where did you go?

'Cause strategically you could go anywhere in terms of where like you could even let it just kind of word of mouth and organically, or you could kind of choose verticals. Like how did you approach that side of things?

Yeah, so we've pretty much grown purely organically and word of mouth. And like the classic advice that we would get is like find a niche and just build Delphi for fitness coaches or just build Delphi for doctors. But I never really saw it that way because it's more of like a horizontal platform shift, like a new form of media, you know, Beehive or Substack, you could have CEOs writing blogs, you could have doctors writing blogs, you can have coaches writing blogs. It's just a new way of scaling your thoughts. So we had doctors, we had authors, we had coaches, celebrities, business owners. The first use case we were just focused on was a new way to scale your expertise instead of a course or a book that you can monetize. But then people started using it for 24/7 audience availability. Some people started using it as a new kind of LinkedIn profile where Hey, you wanna pick my brain? First talk to my digital mind and it's gonna let me know if we should hop on a call. So top of funnel, middle of funnel, and bottom of funnel are the use cases that started to manifest.

That's interesting to go about it that way. Why was it organic that you wanted to go with first and not kinda pour money into this and do that? 'Cause I think you said on a podcast that you wanted to wait until your Series A, which you just raised. Why was it organic from the start? Take me through more of that decision process.

Well, I'm a big believer that if you've built a product that is reliant on paid marketing to scale, then it's not a good enough product. And I think there are some other entrepreneurs who are like, the product is the marketing. If the product is great, it should market itself. And so I never really wanted to rely on ads and put our fate in the hands of Facebook or Instagram, but rather like I wanted to build a product that I was proud of, that other people were proud of. And like Brendan Burchard loves it so much that he'll tell everyone about it.

Yeah, you seem to have cracked that already. And I guess, yeah, just like when you had that moment of like OpenAI with ChatGPT, people were like, They don't have to do paid growth to grow to grow the ChatGPT. Obviously, people are seeing it and telling their friends like this is ridiculously amazing.

Like that's what yeah exactly.

We don't have ChatGPT growth yet, but one day yeah or we're just we're very close. I'm sure. Tell me more about the early days. Even like okay with this idea because I don't want to gloss over that. As most early stage founders here, even raising that pre-seed with this idea and like the what you wanted to build with this. Obviously, you were working with Keith Rabois before. You got a company that you sold within a year, which. Friday. Take me through raising that first round, kind of getting the initial, like, that capital to build this in the first place.

Yeah, I mean, another objective of mine at Open Store was I wanted to raise from Keith. I think he's a really, really smart guy. And so Sam and I worked our asses off at Open Store. And so by the time we went to Keith, he kind of knew that we were really hard workers. And that's the kind of founder that he likes, I guess. So yeah, Keith bet on us because of us. I don't think he actually believed in the idea that much. Maybe he did a little bit. But then everyone else was like— bizarre. Our raise went pre- and post-launch of ChatGPT. So pre-launch of ChatGPT, everyone was like, what is this? I don't get it. This doesn't make sense. And then like a couple days after ChatGPT, they were like, oh my God, this makes a lot of sense. And then a week after ChatGPT, when Character AI launched, they're like, oh, you guys have no shot. You don't own the model foundation. You don't own like the The layers, you don't have no research engineers. So it's just like the opinion just changed so fast over that month.

That had to be a whirlwind, I'm sure, from that time period. Yeah. Well, one thing too I'm curious about is with the product itself. Obviously we see with a lot of these different AI products, they have a simple interface. It's just like a chat interface. Was that always kind of what you envisioned this being and to keep it super simple in that way? I know you didn't do like, you're not trying to do the video side of it, at least in the early days, the video side of it, like having this like Replicating me as, as an AI thing. Was that intentional? Always part of the plan to start with that? Like, tell me more of the product itself, especially.

Yeah, I think we wanted to start simple and like the chat interface is not what I think is the final. We're actually working on some new things right now that I think will make it better, but very simple. Can you talk to the person? Can you get suggested questions? Can you call the person? We actually do have video. It's just not on all of the Delphis because the quality isn't that great. The part of the product that I think you probably haven't seen, which is the majority of the product, is actually the studio experience for uploading your data, connecting your social media, your Notion, uploading your voice, getting insights from conversations, monetizing it, creating actions. There's all these things you can do on the backend to curate the experience.

Tell me more about that in terms of thinking through that side, because obviously you went like B2B. You're like, we need to work with them. We need to get these people who have big audiences. What was the feedback you were getting from people in terms of what they were requesting, things you know you need to provide for them, help that experience as well?

Well, it's a hard balance because this is an entirely new thing. So a lot of people actually didn't know what they wanted, and we just had to think from first principles. In a world where you could be in multiple places at a time, what would you want to know? You would maybe want to know when someone interesting is talking to it. You'd maybe want to know where there are gaps in data. You'd maybe want to know, like, what's resonating with people, you would want it to work on any platform. It shouldn't be a walled garden. It should work on Zoom, SMS, Slack, Discord, Telegram. Maybe you'd wanna monetize it. You'd wanna have a CRM of everyone talking to it. So I just think we had that initial thought process in this world, what would people want? And then we kind of intuition first, data second.

Yeah, 'cause it's an entirely new thing, which is then a whole other complexity of like—

No data. There's no data to base it on.

Yeah, we're building the first version of this, which is so interesting. To that point, like, what is the— what are the complexities of that in terms of actually building this? Like, the engineering challenges, anything you can talk to in terms of actually building this product and your team building this product? Because within AI, there's so much to it. I'm curious on that side of things as well.

It's very difficult. It's difficult on 2 fronts. One, like the mind architecture. How can you represent someone's worldview at any point in time? And the thing about a mind is that, you know, who you are changes based on who you're talking to and what your objective is. Not everyone has a ton of data. Your tone and your beliefs change over time. It's temporal in nature.

Yeah.

And then we want to be able to capture your reasoning so we can be predictive of what you might say in new situations. That way, if I say, hey, what would you do in my situation? I'm asking for personalized advice. We need to be able to capture that reasoning. So that's been difficult. And then the other difficult aspect of it is just UX. Because again, there, there are no patterns for this. Create your digital mind and get insights. And so I'm constantly thinking about like, how can we make this simpler and better? Because we don't want people to get confused by this new thing.

Yeah. And it is, I mean, people can check it out and like delphi.ai to see the experience and like play with it a little bit and get a feel for it and understand like what we're talking about with this. One of the things I'm curious about, 'cause I always love the growth side of things as well. Okay. So organic is what the focus was initially. You wanna like product-led growth. Like are people sharing? Is it good enough product for people to actually share and word of mouth and all of that? But you just raised a Series A led by Sequoia, $16 million. How are you thinking about growth today now that you have a little more capital? You've had more experience with this company in 2 and a half years. Tell me more about that.

To be honest, most of that capital is for hiring. I'm still very bullish on word of mouth and PLG and organic. I think we're thinking a lot about ways— because this product, even as we're talking about it, there are probably listeners that are like, what is this? I don't understand this product.

Yeah.

Hard to conceptualize. And so we're trying to make it easier for people to like— as soon as someone shares a demo of them calling themselves, it's like, oh shit, that's really cool. I get it. And so I think shareable assets is something that we're really focused on right now, just allowing our customers to share more broadly, and so people can understand what they're doing. And we do a lot of— one really successful kind of growth motion, which has been finding a central node in a field of expertise, making them an evangelist by making them super happy with the product. And then they invite us to give a talk at their mastermind group or their conference, which gives us access to 500 more people in that space. And that has been a super successful motion. When we launch our consumer platform later this year, there's going to be a lot of network effects, which again, I don't want to be a company that relies on paid ads. I'm hoping that we can just grow really never relying on paid ads.

Yeah, no, that makes sense. Especially given the product that you have and like how you've gone about it. And because it's like this B2B thing and it's like marketplace thing that you're gonna have those people that have those notes. To that point with in-person stuff, how much are you doing in-person versus online in terms of growing, getting the word out? Like people always think of it as like digital right away, but like you just mentioned the conference thing. Like that is a lot of potential customers right there.

So many. Like I've been, I've traveled so much this year. I've probably spoken at like 20 different conferences this year, maybe more. Yeah, I think the in-person is good for Delphi specifically because it's such a weird idea. Some of these people are like anxious about the fact of, you know, what if it says something wrong? Can I trust this company? Can I trust this guy? So I think it's actually been very high leverage for me to go out there and show that I'm not a psychopath. You know, you can trust Delphi. Yeah, so that's that's been very busy.

Like we think he's not AI, but then we see him in person like okay okay he's not AI.

Is he a person? Who am I talking to?

Well right well right well it's funny because I'm hosting tons of events for the LA Grime like LA specific like newsletter media company all this and like Pete there's demand for in person is so strong and companies want to be in person like in terms of trying like trying to reach customers in person and having that relationship being built through in person which is super fascinating like the sponsor dollars and everything. And also then people like growing through that in terms of like growing their products and stuff. But then one other thing with that is you have like an in-person work culture. And you mentioned Tyler Denk earlier. I know you went on, I think, the retreat, one of the retreats with him, and they're very much remote, like completely down with that. Tell me about the decision. Like you want in-person. Tell me more about that.

I mean, it's a combination of selfishly, like I like being around people and I like being remote makes me really depressed. But also we're innovating on so many different fronts and just like the ability to just like riff in real time, like, oh, the mind architecture. We gotta change this, we gotta change that. I think I felt strongly about it. Keith at Open Store obviously felt strongly about it. And honestly, in the end, like, the most important thing about your company is your people and the relationships. Like, it's— the product is great and the company is based, but the people is what makes the experience and the journey, like, worthwhile. And so we moved to San Francisco, and I think every single person in the company has moved to San Francisco for the company.

Hmm.

And I almost sometimes think that I should exclusively hire outside of San Francisco to maintain that culture of people who have upended their lives and moved to a new city.

They're willing to take that risk and do it because they care that strongly about the company.

Yeah, that's what Sam and I did for Open Store. The Open Store culture was very similar. Everyone was fully bought in.

Well, to that point, how are you thinking about the hiring side of it, building this team? Because with your last company, you sold within a year. So it was like 8 months or 6 months. So you never had time to build a massive team or anything. How are you thinking about that this time around with obviously having so much traction, capital raise, etc., building this team for Delphi?

Yeah, it's funny. Our first hire was a 17-year-old kid from Belgium who had pretty much no work experience. Our third hire was our founding backend engineer, had no professional engineering experience before, and those are some of the best people in the company now. I mean, they've grown a lot. So I think we, we balance hiring experienced people and then young people who have very high slope. I think balancing those two is is both rewarding and it pays off as well in the end. I'm spending most of my days in interviews right now because I think the people are the most important thing. We want to keep a lean team. Like, I don't want to become a 100-person company. I don't think you need to these days, but we'll see.

Okay, we can't just gloss over those 2 things, those 2 people you mentioned. Like, okay, how does that happen? How are you vetting them? They just do good work? Tell me more about those because that's fascinating.

Yeah, so Mateen, 17 years old, he reaches out to me on Twitter and he's like, bro, I love what you're doing. I was building something similar. And I was like, okay, leave me alone. And then he just kept sending me designs and I was like, Sam, should we just give this guy a chance? And so we flew him out to Miami. He slept on my couch, a 17-year-old kid. And he's just working out of the Founders Fund office with us. He's starstruck with all the Founders Fund partners and he's just crushing it. Really incredible kid, had read so many books and just was so so hungry, had great taste in design, and we decided to hire him as our first designer. And he's designed pretty much everything you see in the product. Some people will tell us, like, Delphi, wow, your design team is really cracked. You have really great design. I'm like, it's just Mateen.

He's crushing it for us. Talent's everywhere, man.

I mean, he's great. And Alvin had reached out. He was working at 20VC. as like an analyst or helping Harry. And he was like, hey, I was building something similar called Parrot. Had never really been an engineer at a company before. And he was applying for the founding backend engineer role. And he just like created a presentation of like why he should work at Delphi. And I loved his energy and I was just like, you know what? Yeah, let's do it.

I love that. It's never so like smooth. There's like different— so many different ways to hire and find people. There's talent everywhere, like we said, like One thing is too with that, I mean, you mentioned the design thing. You needed to be buttoned up with that. You needed to have that be very professional in that capacity. One other thing I heard about, like very careful with your messaging around what this is like not saying like clone or chatbot or whatever. Like you're being very deliberate. Tell me more of that and the messaging and like that side of Delphi to get the word across.

Yeah, I think for the wording, like Our whole thing is that when AI is abundant, what becomes precious and special is humanity. It's the scarce resource. And I agree with you, in-person events are going to be even more premium, I think, as AI gets more popular. And so calling a digital mind a chatbot, I almost think is offensive to the person who is scaling themselves. And a clone is just dystopian. We used to use the word clone. Clone yourself was our tagline. everyone immediately thinks of Black Mirror. And I think there's actually like a physical cloning company that bought clone.com. So we're like, you know what, that's really not going to be our space. So instead, what are we? What is it that makes Delphi special? Not necessarily like the voice or the video, but— or that's part of it, but like your mind. Like the most unique thing about you is your mind. And so Delphi enables digital minds and it is a Delphi. People call it Ask My Delphi.

Hmm.

Another thing with that, with that messaging side of it, you mentioned you're just basically in meetings all day with like Hiring, you're trying to build out this team for Delphi. What is the messaging to them? Because especially now, in terms of joining AI companies and stuff, there's a lot of opportunities out there, like competitive market. We see Zuck just throwing hundreds of millions after people. Like, tell me that messaging to potential employees and like, what are you getting? How are you getting them to join Delphi?

I think Delphi is a very different company from most of the AI companies out there because most AI companies right now are very focused on like AI automating efficiencies and productivity, where we didn't start the company because of AI opportunism. We wanted to solve a problem, and we're really focused more on preserving humanity instead of automating it, augmenting humanity. And for some people, that really resonates, and I think those kinds of people are the ones that usually do well. I think half the people at Delphi were working on a similar idea before this, and it's a really weird idea. There's a lot of people who still think it doesn't make any sense. And so those people, when you find them, you're like, all right, You should work here.

They all get it. Yeah, they already understand what you're trying to do, basically.

Yeah.

That makes sense. Where does it go from here in terms of raising that capital? You said a lot of it's going to people. What's the next phase in terms of getting this to the next level? Because obviously you've grown a bit in the last 2 and a half years since you started. Where does it go from here? Like, where's the next— what's the next kind of phases for this?

Yeah, I think it's 3 things. One is continue to improve the value someone gets from having a digital mind, like the insights and just simplifying the experience. 2 is lowering the barrier to entry to creating a digital mind. So we haven't launched our free tier yet. We're planning to launch it soon. In order to launch it, we need a way to give people a really good experience, even if they don't have a blog or a YouTube channel or a podcast. And so we're working on a solution to that. How can you create a high-quality digital mind without any data? So we're working on something called Interview Mode, which I can talk about in like a month.

Yeah, that's all good.

The next phase of that is enabling discoverability. Like, how can you find people based on their expertise? You know, on LinkedIn you can find people based on where they've worked or, you know, what school they went to, but often the best relationships, whether it be someone you want to learn from or someone you want to work with, is based on similarity of thought and like their values and their experiences.

Interesting. One thing too with this product, I mean, I saw like from reading the website and everything else and like seeing like How people are interacting with this itself in terms of like the people who are creating. So there's them, there's different tiers of the product. And obviously this could change when people listen to this like years later, but there's like tiers, tiers that allow people to, for instance, I'm trying to think of like how to word this. They essentially get to see the data in terms of like how many people are, what people are searching for or how they're interacting. I mean, like how they're interacting with their chatbot. What have you guys seen from that side of things in terms of interactions? Like how often people are using this, like how they're asking things, like anything around that side of it that helps you improve the product and also on both sides of it, really?

Yeah, I think it's interesting because Character AI, their number one metric is engagement. And so they're building models to keep people talking to these characters for as long as possible. I don't think that's the metric that we want to optimize for. One, because I'm not trying to build a product that thrives off of addiction. But two, there are some people who use Delphi as a personal website replacement. And if it just answers a couple questions a week, that's like a success for them. Versus like someone like Matthew Hussey, people will call his Delphi for like 6 hours at a time.

Yeah.

And like come back. And, you know, we've seen people whose first experience with the product is through voice end up coming back a lot more. So I think there's a good stat there, but we're trying to think of other ways to measure value gained by the consumer, which may not be Directly proportional to like number of messages in a conversation. I think how often they come back is a better measure than how long the conversation is.

Hmm. That's true. Yeah. Tricky to figure out cuz yeah, you don't actually care if they're engaging a ton in a single conversation, but do they continually come back to it? Like, look at my, I'm thinking of my own behavior, ChatGPT, like how much I go back to it for everything. Oh, let's ask about this. Ask about this. Like, what about input this? Like that behavior. It's not that I have one long session. But when I, yeah, I think of something, it's always going back to this. I guess imagine it's like the same thing with Delphi. It's like, oh yeah, I want to, like, I need to be in Matthew Hussey. I need Matthew Hussey's perspective on this, like every time, or whoever else, like, you know, Arnold Schwarzenegger for health stuff. You would continually just go back to Delphi is the idea.

Yeah. Another way to think about it is, you know, there are some YouTube videos that you watch until the end and there's some YouTube videos that you drop off initially. And like, the fact is some people just aren't interesting. Like it's, it's, it's a feature and not a bug because we don't want to optimize. We don't want to set the metric that every single person's Delphi should get a ton of engagement. You should be an interesting person and as a result it will get a ton of engagement.

That makes sense. One thing I'm really super curious about is like you bootstrap your first company, you sell that in a relatively short amount of time. Now you're going the venture-backed route. You have a co-founder, you've raised, you've raised capital and everything. How are you learning, growing, developing? Because you're building this company that helps people do that through talking with these different people through Delphi. How are you personally as a CEO, like as a founder of this company, how are you learning, growing, navigating things? You have incredible investors. I'm really curious about that side of things as well.

Well, I think I'm really lucky, one, to work with incredible investors that I get to learn from. And we invite our investors and some great people to the office to give talks to the team, and that's really fun. I also just get to work with some amazing people like Brendon. I used to be a horrible public speaker, and Brendon taught me how to be a good public speaker.

Really?

Yes, I used to be nervous. And so it's interesting that I get to work with these experts and almost learn from them and adopt their skill sets, which is super cool. But otherwise, I barely have time to read anymore, so I'm very much like dive into the weeds to learn. Like, okay, I need to learn how to be a better brander. Let's just Do the rebrand and see, and see what happens. And I'll, I'll trial and error my way to the top and I'll, I'll talk to a couple Delphis along the way to get different perspective.

Yeah. How are you using AI in general? Like tools you're using, think, how are you using it in the company? Like I'm super curious about that. Anyone who's building an AI company, I'm always curious about like what do they know? What are they, what are they using? So tell me more about that.

I can't say I've been doing anything crazy. I mean like Perplexity when I want to get research on someone before I meet them. Claude has been amazing for if you're sending difficult emails, whether it's negotiations in a fundraise or sending an email to a customer who's upset. I always use Claude as an AI translation tool. And I don't mean language translation, but how someone perceives your language might be different than what you actually mean. And sometimes that nuance gets lost between emails. And that's actually been super helpful in resolving conflicts that sometimes appear over email. But yeah, otherwise I think not much. I don't really use many other tools.

Yeah. I mean, there's a few essentials that you can get a lot of use out of and be okay pretty much. The other thing I'm curious about too is you mentioned, I don't know if we talked about this right before recording, but this company started in Miami. You were in Miami before. You're in SF right now. Tell me about location-wise, like the idea, it was super hot back in like 2021. People talking about like moving different locations and like you can work from anywhere. It's like, well, you can be wherever. Like it's changed a bit. Tell me how you think about that location-wise. And I know you love in-person culture. You said talent is in SF. That's a big reason. But just tell me how you think about that location-wise, moving to different places for your company, building in that city even. Like what does that mean for you? Like I would just love to hear more about that.

Yeah, I loved Miami. The quality of life was amazing, but at a certain point we were thinking, If I am someone, and this wasn't even because of the current talent in San Francisco, we were thinking, if you're a young and ambitious person, what are the things you're optimizing for when you think about what you want to do? You want to go somewhere where you can be the best of the best. If you want to be the best actor in the world, you're moving to LA. If you want to be the best trader in the world, you're moving to New York. If you want to be the best politician, you're moving to DC. And it's just like San Francisco will be the only place where we can at least set ourselves up for success for being in a place where the best young people may want to end up in.

Yeah, it is something. I mean, there's a lot of Paul Graham essays you could dive into, one being like cities and like how they tell you a story and everything, which I love, and like the ambition and what they're speaking to you and says about it. Like, I mean, I think I'm in LA, and I think a lot about building LA as like a startup hub. Obviously SF is like number one by far, but thinking about how you develop that in a city and like what are the components of that. Is kind of a fascinating problem to try to solve because it's like, okay, do you have enough engineering talent? There's ambition here, of course. You need enough wins where the money and capital can flow back into earlier stage companies. Like there's so many of these different components that go back into an ecosystem to like figure out how to grow something. That to me is always like the most interesting thing because like it's not SF, but how do you build it somewhere else? Or how do you build it in Austin, Texas or whatever? Like it's super interesting.

I think Miami was close. I think the 3 things that you need that it didn't have, one is like a school. Or like 1 or 2 really good engineering schools. Another is a couple winners where like the alumni of those winners are starting companies, investing, and, you know, maybe Snapchat is that in LA. And then the last is multiple parties developing community. So the problem with Miami is that there was really only 1 group planning parties and communities, like the Founders Fund and all of that, but you need like multiple pockets versus just like 1 pocket.

Yeah, I think that makes a ton of sense. Even seeing SF, I go up there a lot and I interviewed Dave from HF Zero. They have the 20,000 square foot house for their residency, which can be a center for a lot of things. And you can see how much, so much density. But there's so many different community builders to your point. If I look at there, I can look at Jonathan from Brex. I can look at Michelle from Headline. I can look at all these different ones that are there. And then obviously you have like the HF Zeros of the world, other residencies as well, like The Residency.

Yeah.

there's so many, which then helps the ecosystem. And I remember even like when I was there talking to Dave, it was like, oh yeah, they had, they heard they'd be like talking about something or building something as a product there. They'd have a research paper in like, you know, Palo Alto that like just got published. They'd have that person come to AGF Zero the next day and like build it that night. Something off of that new research. That's insane. That's so cool to be able to build something like that.

Yeah. No, there's definitely a lot of serendipity in the city and I lived here in 2020 and I hated it, but I will say it's, definitely a lot different. The new mayor, there's new like bars, new restaurants. There's definitely like an increase in energy that I can feel.

Tell me, with Delphi, you mentioned a little bit about kind of where this is going. What is the grand vision ambition for this? Just like one, you personally, but also with the company, because like everyone has a different, different idea of what they want to be. You've seen, like you said, your grandpa had like 30,000 employees in Iran. Like, that's insane. You've seen that, like you've heard of that. Like, what do you want this to be like? What's your ambition for that? Like, I'm super curious about that with Delphi.

Yeah, I think there's a couple facets that I hope for Delphi. One, you know, Bloom's 2 Sigma is this study that shows that anyone who has like a mentor is better off by like 2 standard deviations, like a private tutor. And I hope that the effects that enabling mentorship with some of the best that it has on like young people does have that same effect. Like, you know, any young person doesn't feel like they're alone or they can talk to whoever they want to talk to. And it's just like a more engaging way of learning. Because I do feel like, and I've heard this from our customers, people are reading books less and finishing courses less because unfortunately TikTok and Reels are just depleting attention spans. And so hopefully there is a better way to learn that is engaging and gets people excited and inspired. So that's one aspect of it. Another aspect of it is if a company like Delphi were to exist, it should be around 400 years from now, like long after I'm gone. Every human being their legacy can be preserved on this platform like the Library of Alexandria. So I would like it to be a company that outlives me. And then number 3 is the people-finding aspect, like finding the best person in the world for whatever you want based on similarity of thought, which I think leads to the best kinds of relationships.

Well, I mean, to that point with this, do you see that as some kind of like Moat is always hard with companies, like to talk about moats and everything. And it's always like, it's not very strong, but an example, a podcast coming out, it will be live by the time this is out. Obviously I talked to the founder of Bordi, uh, Andrew D'Souza, and they have done at this time, I interviewed him like maybe 40, I think 50,000 conversations with people, voice recordings. They have that. They can find people like that moat. If you wanna talk about moat, it's pretty impressive. And like, you can't just like instantly replicate that. How do you think about that for DeFi? Because if you're saying this, you have all these people on the platform, like they don't have an incentive necessarily to be their AI multiple places necessarily, unless the discovery is better somewhere else. How do you think about that piece of it and like needing to go fast because of that? Like, tell me more about that.

Yeah, I mean, I think the best moats are product moats in the end. You know, when we were raising our seed round, everyone was like, what's your moat? And I would always— I remember Keith and I was like, Keith, all these people are asking us what our moat is, and we're not training our models. And he was like, anyone who asks you what your moat is this early, you should never talk to them again. They're not serious investors. And I think if you just look at the history of billion-dollar consumer companies, none of them really had a moat early on. They shipped really fast. They really cared about the details of the product, details that most other people don't care about.

Yeah.

And I think for us, I could say our mind architecture is very unique, and it really captures someone's worldview. That's true. Can it be replicated? Maybe. I could say that people are only going to build one digital version of themselves, and they're going to want it to work everywhere. So the more people we get on Delphi, there's almost like switching costs to moving your data to another platform. And then there are the network effects of the consumer platform. So those are like potential moats. But again, I really try not to think about moats and really just try to think about like, how can I make the best product that we— that could possibly Possibly be out there.

Yeah. And your culture is such that you're obviously working very hard to do that and shipping pretty quickly. And I've seen some of that over the last even couple weeks of like stuff you've done, which is interesting. Where's the best place for people to reach out, connect with you, learn more about Delphi? Like you're hiring a bunch as well. Tell me more about that.

Yeah, the best way to learn about Delphi is to go talk to my Delphi, delphi.ai/dara. And then I post a lot on Twitter. Dara Lajj and LinkedIn about company updates. And yeah, my, my Delphi will let me know if we should hop on a call, so I don't need to give them my email.

I love that. I love that. And I've seen, yeah, on Twitter, LinkedIn and stuff like as well. Founders who are creating content like that, it makes so, so much sense for the organic, especially given what you're building is like anyone can kind of search it and use it. It's so relevant, which makes a ton of sense. Yeah, totally. Dara, thank you so much for the time today. I really appreciate it.

Thanks so much, Justin.