Interview with Stephen Stearman | Episode #26

What does it take to build an AI product that clinicians actually trust? We invite Synth.ai CEO Steven Steerman to break down how his team turns hours of psychological report writing into minutes, while navigating the hard realities of HIPAA, SOC 2, GDPR, and the fragile trust that keeps practitioners coming back. Steven shares the inside story of a venture studio spin-out, the early MVP stitched together on foundation models, and the evolution to specialized pipelines that route scans, handwriting, and docx through the right OCR and LLM stack. The tech is smart, but the lesson is sharper: purpose-built beats general when the workflow is high stakes.

We dive into the part few founders talk about openly: compliance as product. From business associate obligations to breach notification and international requirements, Steven explains how they are accelerating audits without compromising sensitive mental health data. He also gets candid about the tradeoffs that keep him up at night, pushing features to win new customers versus deepening polish to retain trust, all with a lean team and a real burn rate. His north star is disarmingly simple: deliver a magic moment in the first 60 seconds, then keep the experience consistent after week ten.

On go-to-market, we unpack a surprise: paid ads that finally work. The winning angles are honest and human, write reports in minutes, stop filling tables, get weekends back, paired with tight creative that speaks to small-practice psychologists who buy like consumers. We also address the elephant in the room: LLMs as both competitor and substitute. If a savvy clinician can wire prompts in ChatGPT, Synth must be more compliant, more convenient, and unmistakably tailored to the clinical workflow.

We close by zooming out. AI wrappers on rented land face platform risk, especially as big labs watch popular use cases. The counter is to own the workflow, build domain-specific pipelines, and invest in brand, community, and narrative, now the real moats. If you care about AI in healthcare, product-market fit, and the delicate craft of trust in the age of compliance, this is your playbook. If you enjoyed the episode, follow the show, share it with a friend, and leave a review. What part of the build would you double down on next?

Chapters

• 0:10 Hosts Introduce Episode And Guests
• 1:53 Clubs, Crafts, Slime Businesses, And Nostalgia
• 9:48 Trips, Cities, Cost Of Living, And Lifestyle
• 24:29 Transition To Interview With Steven
• 24:50 Steven’s Background And Venture Studio Model
• 29:45 Synth’s Problem, Traction, And Burn Math
• 32:27 Compliance Deep Dive: SOC 2, GDPR, HIPAA
• 38:32 Why AI In Psychology: Pain Points And Impact
• 40:33 Under The Hood: Models, Pipelines, And OCR
• 45:14 Trust, Hallucinations, And Damage Control
• 48:16 Paid Ads That Finally Work And Creative
• 53:27 UX Magic Moments And Voice Agents
• 1:00:02 Adoption Curve, Prosumers, And AI Sentiment
• 1:04:16 Founder Lessons: Speed Vs Beauty, Focus Vs Features
• 1:07:12 Post-Interview Reflections
• 1:10:03 Rebrand Shortlist And Naming Debate
• 1:19:08 AI Wrappers, Moats, And Rented Land
• 1:24:07 Are AI Platforms Spying On You
• 1:28:06 A16Z Media Fellowship And Community As Moat
• 1:33:20 Tease For Next Guest And Closing

Transcript

SPEAKER_02: 0:10

And we're back, episode 26 of the AI Code Central Podcast. This is your boy Charlie. We got Connor Douglas. He's a paid ads marketer expert. And we've got Mallory Lore. Let's go. Community and brand experts. Again, episode 26. We've got a good interview with our old uh our old pal, Steven Stirman. He's a CEO of synth.ai. He's a beautiful animal. He's a legend. Yeah. So he so for context, uh, that is the project that Connor and I met on. I was doing the SEO. My agency was doing the SEO. Connor's agency was doing like uh emails, analytics. Analytics, yeah. And um, I I I do remember that. And uh that was back in 2021. So that's how we met. And he transitioned into uh a pretty cool AI company that solves a really big um problem in the uh in his in his niche. So um good to chat with him. That was about I don't know, 45-minute chat, would you say?

SPEAKER_03: 1:14

Yeah, like 40 45 minutes again. Just covered a really interesting topic, at least to me, in terms of kind of that health space when it comes to leveraging AI and super insightful. And like you said too, Charlie, it's just he's covering, I mean, he's tapping into a market that I just don't think anybody else is really at this point. So it sounds promising for him. I'm excited to uh to see how he grows as time goes on.

SPEAKER_00: 1:38

Yeah, and I think I think for me, just trying to understand like how you really build trust in like the era of compliance with AI, like what is that like? So I'm curious to learn more about that.

SPEAKER_02: 1:50

Yeah, yeah, for sure. That was uh that was a good topic. So before we get into that interview, uh Mallory, what's been new this week?

SPEAKER_00: 2:00

What's new in Mallory's world? Okay, um, let's see. So this past weekend, um, I love to make clubs. So I have a brunch club, I have a craft club that I brunch club.

SPEAKER_02: 2:11

Yes, yeah, brunch club. Spot each month, right?

SPEAKER_00: 2:14

So a different spot each month. I think I said this in the last pod. It's been running for two years. It's it happen it's on our calendar every month, it's recurring. So that way, like, look, if you can't make it this month, come on over to next month. Like, I just instead of planning all these things all the time, I just wanted to like make it a consistent thing. Um, and so now I have started a new club called Craft Club. And this week we worked on these little clay polymer uh magnets. Um, I'm obsessed with craft TikTok. This is an avocado, although it may look like something.

SPEAKER_03: 2:47

That's cool.

SPEAKER_00: 2:48

This is a little cookie. So, anyways, obsessed with crafts, right?

SPEAKER_03: 2:52

These look like things that they sell, like in like I have a store by me. It's it's called Lockwood. Like everybody has their little gift shops and stuff, right? Like this looks exactly like things they would sell in there. That's so cool. I love that.

SPEAKER_02: 3:04

Did you make mention? Uh was it what well it had to be two episodes ago where like it looks like one of those things, like you push the thing out, like the little poop comes out of the yeah, yeah, the little poo thing, or you squeeze in the eyeballs, like blasting like yes, the slime, the slime.

SPEAKER_03: 3:18

Yes, the slime things, like those are cool. Oh, that's fun. I love the trinket.

SPEAKER_00: 3:22

Also, yeah, I love the trinket stuff. The slime stuff's interesting because that whole market is fascinating to me. Like people like overnight making millions and millions and millions of dollars over launching slime companies, which is topic for another day.

SPEAKER_02: 3:35

That's a thing.

SPEAKER_00: 3:35

But um, that is a thing.

SPEAKER_03: 3:37

Uh TikTok, you can make slime anything though.

SPEAKER_00: 3:41

Slime spray like back in 2021 or 2020, I felt like it was the rise of slime content. It was like ASMR, like, watch me make this kind of slime and this kind of slime, and then people wanted to buy it, and so people were making businesses out of it, like overnight, and fairly successful businesses on TikTok.

SPEAKER_03: 4:00

So it's like a sensory thing for a lot of people, right? They like that. So exactly that.

SPEAKER_00: 4:05

Um, but what else? Uh honestly, the rest of my weekend's pretty chill. I'm starting to watch Harry Potter because it's a tradition for me every time this year, as a the lame ass millennial that I am, to re-watch the Harry Potter movies. So I'm also doing that.

SPEAKER_02: 4:19

So uh my my girlfriend loves Harry Potter, loves Harry Potter. I told her a story when we were driving over the weekend that uh when Halo 2 came out, I was the Xbox player. By the way, I was such a loser. I was like, I just I played Halo, I was actually a really good Halo player. I was like nationally ranked like top 10 at one point.

SPEAKER_00: 4:39

That's sick. Why does that make you a loser? That's that's awesome.

SPEAKER_02: 4:42

Well, okay, I guess back then I was a loser. Now I'd be in like a hardcore, like e-game like Hall of Famer, like like retired jersey in the rafters, like a whole ceremony for me, all of that. So Halo 2 is my game. So anyway, maybe it's a Reddit thread. I don't know. So somehow I found out that in uh Harry Potter's Snape kills Dumbledore, and I had no context of what that meant. So before that book was debuted, I would go out and run around, I would jump in the Jeep and lay on the horn. Snape kills Dumbledore, Snape killed Dumbledore. I was that guy. Charlie, I know.

SPEAKER_00: 5:18

All right, well, this is the last day of my podcast. So this podcast of uh but I I told my girlfriend she was like, Oh my god, you did not.

SPEAKER_02: 5:27

She like starts like shaking like like that. She goes, No, you don't know. There they were friends that like nobody knew who who Snape was, or like Dumble, maybe it was Dumbledore. I don't know. She like told me the whole thing. I was like, okay, kind of.

SPEAKER_00: 5:39

You need to go watch the movies. So you've never seen them at Ultra.

SPEAKER_02: 5:42

I've I've never seen the movies, I've never read the book, dude.

SPEAKER_03: 5:44

They're great movies. Like, I'm not like a dot, I'm like a Lord of the Rings guy. Like, I love those. Like, I could watch those non-stop like every day. But Harry Potter's fantastic. Like, you have to watch it. I've never okay. I I've never got into it. Other than other than Mallory, they have um, I just got uh an ad for it because I'm bit you know, New York City, big on Broadway. You guys have Broadway down in in SF as well, so you know by you. Um, but Tom Felton, the guy who played Draco Malfoy, is now playing it in New York City. He's playing himself because it the play takes place when they're all older. And we've seen it, my wife and I have seen it before, but we're gonna go back just because I mean Tom Felton, he's in it. It looks awesome.

SPEAKER_00: 6:27

Oh, I want to see it. Okay, that sounds really cool.

SPEAKER_03: 6:29

Have you ever seen the the play?

SPEAKER_00: 6:31

I haven't seen the play.

SPEAKER_03: 6:32

Oh, it's really good. If you have have no, but I need to.

SPEAKER_00: 6:35

I need to go for it.

SPEAKER_03: 6:37

It's awesome.

SPEAKER_00: 6:38

And Charlie, you need to watch the frickin' movies.

SPEAKER_02: 6:40

Okay, that's really bad.

SPEAKER_00: 6:42

When you read that book, wait, and your girlfriend likes Harry Potter. How is she how is she not had you watch these movies?

SPEAKER_02: 6:50

Right, like you're not being forced to. Yeah, she she was trying to I see negotiations. I see. I I think we settled on like a hurricanes game, you know, because she's from North North Carolina, so we we settled on like a hurricanes game, but okay, I I will watch one.

SPEAKER_00: 7:04

Um read the book that you sent me, okay?

SPEAKER_02: 7:06

Okay, so when all right, at the same time, you read the book, I'll watch the yeah, there it is. When without pitching manifesto, yeah, great.

SPEAKER_00: 7:13

This is the book that Charlie sent me last week to prep for one of my potential uh client calls and to lock lock it down. So um I haven't read it yet, and he gave me a really hard time about it, but I will read it.

SPEAKER_02: 7:25

Um, had to take out a pair of scissors and like open the box. Like, I sent it to you like five days ago. Like the call was on Friday. Made sure it got there by Thursday.

SPEAKER_00: 7:37

Oh, anyways. Anyways, that was basically I don't oh, what else? Um, launched my Substack last week, which was really exciting. Um, if you're curious to follow, uh, it'll be a place where I post like all things brand, community marketing, um, sort of the intersection of AI and culture. I'm about to come out with a piece uh this Thursday that focuses on the fact that the world needs less brands and more communities. Um and I'm gonna dig into that. And so really excited about that. Um I'm trying to post every Thursday. That's my goal. Yep, yep, yep. Um, just try to try and be really consistent there. And then I've been vibe coding and I've been enjoying vibe coding my portfolio site, and I'll probably end up vibe coding our agency site just to like start somewhere simple. But um I've been having fun doing that too.

SPEAKER_03: 8:29

So isn't it crazy how you can kind of just fall into like you start the vibe coding stuff, really just leveraging AF for anything? Like I'm using it for script writing for two big projects, and I I don't even hate it sometimes. I actually thoroughly enjoy it because it's just so cool to just go back and forth. And when you finally get like the output you want and you get it right, you're like, yes, like the the feeling of just accomplishment is so great when I'm using stuff like that.

SPEAKER_00: 8:53

But anyway, no, no, it is for me too. Um, and so I'll like go in ChatGPT and I'll have it like give me, which I could probably just do in bolt. I don't know why I'm doing this, but I go into Chat GPT and then I share the output and prompt with bolt. I'm like, I want these changes, help me revise the prompt and then going back into bolt and then giving it this like every single time. It's not like a just do that. It's like a very in-depth detailed detail.

SPEAKER_02: 9:19

Did you start with a PRD? Or is it just like uh here's context of like here's who I am, here's the agency that I want, and then you kind of spit it out.

SPEAKER_00: 9:27

I mean both. It was like here's like the one pager that I created about my agency, and here's like exactly what I would like out of the site. Like, um, so it was kind of a mix of both, giving like some references and outlining what the PRD would look like. Um, so yes, and that's been my my time over the last week, guys. What about you, Connor?

SPEAKER_03: 9:48

Um yeah, well, I haven't I haven't been on with you guys. This is I'm back after a a week off uh hiatus or vacation.

SPEAKER_02: 9:55

By the way, we we put sorry to cut you off. We put in the header of the uh the show outline. I was like, Oh, I guess Connor's here too. Like, yeah, thanks a lot, guys.

SPEAKER_00: 10:06

The disrespect.

SPEAKER_03: 10:08

Okay, I'm back. Yeah, and very, very involved this episode. Okay, so so back off. No, anyway. Um, last two weeks have been pretty good. Uh again, I missed the last episode. I was up in Vermont. Um, my wife and I, again, our whole thing is we want to see all 50 states. Vermont was one of the easiest ones to get to. Quick drive, you know, only about five hours for us. So went up. It was nice, you know, beautiful. It actually, so we went to two places, went to Stowe and went to Burlington. Burlington's like the bigger city. Stowe is more so. Think about like um picture a movie where they would like maybe like burn witches at the stake. And that's what Stowe reminded me of for some reason. Just like very old-looking churches and homes, still really pretty. Um, but it ended up snowing up there, which was really, really nice. You know, we like the snow and all that stuff. Um, isn't that what the witch project was?

SPEAKER_00: 10:57

No, that's Salem. Salem, uh, okay, right?

SPEAKER_03: 11:01

No, maybe not. I don't know where Blair Witch Project was, but Salem is in Massachusetts. I've never been there. Oh, okay. I would love to go because I've got that we just scared the hell out of me when I was in middle school.

SPEAKER_02: 11:11

Now it's the stupidest freaking movie or something.

SPEAKER_03: 11:13

Yeah, I mean, what's the it was a scary movie where they have the close-up of the camera and someone's like nose is just constantly dripping? Like their buggers are like, yeah, yeah, yeah, yeah.

SPEAKER_02: 11:20

Or the family guy who's like, yeah, something about a map, something about a map, movie's over, everybody looks pissed.

SPEAKER_03: 11:26

Yeah, exactly. But uh, that was a cool trip. And then the last week I've been just like knee deep in creating and customized CRM for two clients. And you know, go high level is custom CRM. It's like they so it's not it's a customized go high level, right? And it's like go high level has like all these APIs you can write and you know, all these connectors and everything, but there's nothing that's like native in terms of what these clients want to integrate. And as much as I kind of push back, I'm like, hey, maybe we go a different route or you do this or that, they're just like, nope, I want to use high level and this is what I want it to do. I'm like, you know, being the agency, you kind of have to appease and say, sure, like we'll try and put it together. And I've just been, I mean, doing it myself because I've tried to hire devs to do like script writing, like front-end script writing, and they want like 125 to 150 an hour. I get it, you got to make it, it's difficult to do. But at that point, it's like, I'm just gonna go over to Claud Code, plug in what I need, ask it, and listen, maybe it takes me an extra hour or two than it would for the the devs to do, but I've done it all myself and I've done it with relative ease where I'm just able to get it accomplished. It's just so much customization that they want that kind of becomes a headache. I mean, Charlie, you would like campaign pods. It's like having to go back and forth and back and forth where I've had to give it a rest. I mean, I think last week alone, when I got back, I got back Tuesday, Wednesday, Thursday, and Friday were like 14 to 16 hour days for me just to get all this stuff done. Which again, listen, that's what you got to do sometimes, right? It's there's there's some bad with the good, but yep. Um, so I've been really, really, really into that. And then yesterday, I'm not a huge like wrestling fan, but my buddy is a really, really big WWE fan. Like, we we love WWE growing up, right? Like, you know, Stone Cold, Steve Walls and the Rocky. Yeah, dude, I was a huge, huge wrestling fan. Because you're like like backyard wrestling and third. Oh, dude, we had a trampoline, we would like throw each other all over the place. Like, like we would get out like a mini ladder, and then my parents would come out and scream, we'd be like, What are you doing? Like, put the ladder away. Like, like hit each other with a ladder, dude. Hit each other with ladders, with chairs, like dumb kid stuff. Oh my god. So my friends maintained his passion for wrestling, you know, and and I, you know, I kind of keep up with it. I just I don't tune into it like that anymore. But um, so John Cena, which I think we all kind of know, John Cena, whether from wrestling or being an actor. Do you guys know who he is?

SPEAKER_00: 13:46

So yeah, he follows me on X because he follows everyone on X. He does what? I don't know. I feel like he does, but he follows me. So we're we're besties, guys. What that's crazy. Does he really follow you on X? He does, I think. Yeah.

SPEAKER_03: 14:00

Oh, that's pretty cool. But it was his very last like wrestling like match or something. So my butt, and it's at Madison Square Garden. So we got these like amazing seats, like maybe like 10 rows back from the ring, a little further. And we just went yesterday and we just had a blast. It was cool because it's nostalgic for me, right? It's kind of like this is what I grew up with. This is how him and I became friends when we were like 12 years old, and now we're you know, we're both 37. So it's just kind of a nice time to spend together and be nostalgic. So that was my weekend. It was a nice getaway after the 14, 16 hours, you know, last week and just kind of taking it easy yesterday at least and enjoying some quality time with friends.

SPEAKER_00: 14:41

So that was my weeks.

SPEAKER_03: 14:44

Did you bring a sign? No, I didn't bring any sign. I I wasn't that committed to it. You know, there's obviously plenty of it's crazy how some of these people go all out. Like, I respect it. It's like Comic Con dressing up. No, no, no, they don't give us any signs or anything. Oh, okay. Um but some people go with signs, some people are like dressed up as like their favorite wrestlers, they're bringing like replica belts, like they're just so into it. And some of these people have like these outfits that are they're insane, like it looks so cool. But I'm just not into it like that. I'm not passionate enough to take the time to do it, but I commend the people who do really, really do.

SPEAKER_00: 15:19

It makes the experience super fun, you know? It does, that's good.

SPEAKER_03: 15:22

It does, but uh, but yeah, that's that's been my past few weeks, so you know, staying busy as usual. But uh Charlie, what's what's going on with you, man?

SPEAKER_02: 15:30

Yeah, I went to uh so I I live in Colorado, ready to move, and girlfriend and I went to uh we buzz over to Raleigh. So it's from DIA from Denver to Raleigh. Um, I don't know, it's three and a half hour, just shy of three and a half hour flight, so not bad. Uh buzzed over there, and we've got a rental. No word of I I have this in my notepad, no word of a lie. We turn on the uh the radio. First, it's already playing. The song it's already playing as we're driving back is Kiss from a Rose by C. Yes.

SPEAKER_00: 16:03

It's a classic. It's a classic.

SPEAKER_02: 16:04

Oh, it's all time banger, all timer, absolutely. And I started laughing. She goes, Oh my god, don't you talk about that on your podcast? I started laughing. Yes, I was like, okay, so yeah.

SPEAKER_03: 16:20

Maybe this is the theme. We can kind of rip it a little bit for central. Like, I don't know, there's probably copyright infringement there, but how long has it been? Isn't that done after a certain amount of time?

SPEAKER_02: 16:31

Oh, is it oh the IP? Okay, yeah, yeah, yeah. Um, yeah, so when that played, I was like, all right, yeah, it's gonna be a good trip. So we got a place in Kerry, uh, it's a suburb of Raleigh. We checked out some places, uh, talked to her, or we met with her friends from college, which is super cool. Um, yeah, Raleigh's a cool place, very, very friendly. It's not that Colorado isn't a friendly spot, but I mean, like, you're in Raleigh, and we are at a barbecue place. Unbelievable barbecue, incredible barbecue. The best down there, man. Oh, I had uh smoked turkey.

SPEAKER_03: 17:04

It was amazing. Although I will say Texas Barbecue, Ranger, like Terry Blacks, like there's just I don't know, it's so good.

SPEAKER_02: 17:13

Like, we we met up with her parents, um, and we went to the Canes game, Hurricanes, they were playing Edmonton, which is pretty cool because the best player in the NHL, well, the two best players play for Edmonton, Con McDavid and uh Leon Dreisidel. And um so her parents were sitting right behind us, we're in front, and I was talking to the guy to the right. I was like, just watch number 97. Like the guy to my right didn't really know that much, so I was kind of telling my daughter. I was like, just watch 97, watch how fast that guy is because he's he's an alien. And um they ended up losing in overtime, Carolina lost, but there's a buzz in that arena. Like they they love their canes hockey, which is pretty cool. Um so that was cool. Yeah, it was cool seeing her parents. Um, you know, her her her mom and I have have banter back and forth, and yeah uh her her dad is super cool. Uh I call him Disco Stew. He's the man. Disco Stew. Yeah, from the Simpsons, yeah. And um, so that that was cool. Uh I have a couple notes here. Got the yeah, first song, and then yeah, trip to Raleigh. Like, Raleigh, I uh to me, like especially driving around at night because I'm so used to Denver, like Denver proper with all the lights, and it it's the city, it's it's a big city. A lot of those houses to me, I just got the feel were haunted, and it was kind of creepy like driving around. Like, I mean, where in the Carolinas? Yeah, yeah. So I I don't know. It was just like I I didn't go in we had like a pretty good sized shed in the backyard, and I wouldn't have reason to go in there during the Airbnb stay, but I didn't because I was afraid it was haunted because it looked oh my god, it looked haunted.

SPEAKER_00: 18:49

And Charlie's afraid of ghosts every day.

SPEAKER_02: 18:51

Oh, terrified of ghosts.

SPEAKER_00: 18:54

Wait, have you ever had a ghost encounter to be scared of?

SPEAKER_02: 18:57

I mean, I don't know that I haven't seen a ghost because uh like maybe they're next to my bed when I was sleeping last night. I don't know.

SPEAKER_03: 19:03

That's a great seen a ghost.

SPEAKER_02: 19:06

I don't know that they weren't next to my bed, but like how do I know that? I don't know. Uh um, so yeah, that that city was cool, great vibe. Um, that was fun. And then yeah, we we saw her friends from college, so but they she's known them for for a long, long time.

SPEAKER_03: 19:22

So is the plan you you're gonna you're thinking about moving there, or you know for sure you you guys are gonna move there? Um like eight 80%.

unknown: 19:29

Wow.

SPEAKER_03: 19:29

That's exciting. Were you looking at places while you were there too? Or was it?

SPEAKER_02: 19:32

Yeah, so we were looking at places. We were driving to like different neighborhoods, we went through like downtown and then like some suburbs, like here's a downtown, and then we went like directly like north and then like to the east, and then we just got kind of the the lay of the land, and um yeah, Denver is a very high-priced uh city for uh not quite like SF or or New York, I guess.

SPEAKER_00: 19:55

Um yeah, I know it's really expensive here. There's no doubt about it.

SPEAKER_03: 19:59

Yeah, I'd say they're probably New York City and SF are probably on par with one another when it comes to cost. It's insane. Yeah, absolutely insane. But you're so close to everything, that's what makes it kind of worth it.

SPEAKER_00: 20:13

I mean, I think that's what I'm paying for. Like what I tell people, because everyone's like, oh, you live in San Francisco, it's so expensive, and it is. Uh, but at the same time, I feel like I'm paying for like a quality of life sort of thing. Like I can go surfing whenever I want, I can go to, you know, down south to Santa Cruz or Monterey, I can go up north and explore the Redwoods, or I could go to Napa and Sonoma and drink wine. I could do a quick road trip to Tahoe and I can go skiing and snowboarding there. Like the access that I have to sort of, yeah, everything that's everything. Exactly. It's just to me, I'm like, it's worth it. Now, if you're someone who's like, you know, I know folks that like come to the Bay Area and they're just here to work and not really sort of enjoy the what the city has to offer, um, which is totally fine if that's what you want to do.

SPEAKER_03: 21:00

But um I think it's like if you're more of like a homebody, like bigger cities aren't really like so, yes, thank you.

SPEAKER_00: 21:07

I was gonna say, like, if you're looking for a nightlife city, this ain't your city. You should live in New York, so there's that. But also, if you're not someone who like likes to hike and enjoy nature and all that good stuff, then like you might not be getting, you know, um, what you really can get out of the Bay Area in in San Francisco. So um, so maybe it's just not worth it if you're not that that's kind of so interesting to hear because Mallory, last time I went to to SF, it it was probably just after COVID had been done.

SPEAKER_03: 21:38

So it was still a little kind of shaky with travel and you know, the but the nightlife there was kind of lacking, and it's like well, I'm big, my wife and I are big on dinners, going out for a cocktail, a show, just something, you know, and it it just like after dark in in SF, it just everything kind of closes early, and I was really shocked by it. But I guess that kind of just validates, you know, that's just the the lifestyle there, right? It's a little bit earlier.

SPEAKER_00: 22:02

And look, like you could come here and definitely have a great dinner and cocktail and go to a show. Like it's possible, but it's not like it's not like the city's known for it, you know what I mean? Like you're gonna get so much more of that, even in like New York, and I would even maybe argue Chicago and LA, but like San Francisco's a little bit of a a sleepy city, and people don't realize there's less than a million people that live here. It's it's a very small city, so it's not anywho. I love it. And I and I when I first came to San Francisco, I hated it. And that's I don't know if we talked about this, but um, yeah, so but it is expensive, no doubt.

SPEAKER_02: 22:38

So we we saw some cool spaces, uh spaces, places, uh, saw cool neighborhoods, uh, people were cool. I told Connor this when we we were chatting on Friday about campaign pilots. Like, holy smokes, there's some slow drivers. Goodness gracious. Uh that's that was an adjustment.

SPEAKER_00: 22:55

So I'm surprised. It's the East Coast. I feel like it would be a little, you know, more aggressive, even if it's not New York, you know what I mean? Like, I don't know why I assumed that.

SPEAKER_03: 23:05

Um I feel like it is kind of like I mean, New York City is I I feel like I can drive anywhere after driving where I drive every day. It's like it's absolutely nuts. But that is a little surprising for down south because I feel like it's like I just feel like they always drive fast down there every time I visit.

SPEAKER_02: 23:21

But um, yeah, other than that, it's a cool spot. I I'm gonna I I will move out of Colorado and um I I'm gonna miss the mountains. I I'm gonna miss you know, just buzzing up an hour and a half west to go downhill, you know, go throw myself down a mountain. I'm gonna miss it.

SPEAKER_03: 23:36

How close are you to the mountains in in Raleigh though, if you go there?

SPEAKER_02: 23:39

It's not the same because I I I was talking to ChatGPT and and some of the guys at the hockey hockey game, and it's like it it's not the same as the Rockies. They're there, but you know, it's it's not the same as like Vail or Breck or Winter Park or Keystone or whatever else. So um you guys will get it. When when when y'all come here and you know we get to places and hammer out some content uh here in the mountains of Colorado, you'll get it. Yeah, there's a vibe, so um so yeah, that was it. That was uh it was it was a good trip. Smooth flights too. And um I was gonna ask what y'all watch on plane flights, but I think we'd save that for for uh a different episode. So um that'll be the next one. Um cool. Let's uh should we throw it over to Steven? Yeah, let's take it off with Steerman. Okay, cool. Here's the interview. And we're here, we're sitting with Steven Steerman, the beautiful animal. He's the CEO of synth.ai. We've actually worked together in many, many years ago. Uh, myself with my old SCO agency and and um Conor with his his uh paid ads agency. So um, Steven, thank you for taking the time and thank you for being on the show.

SPEAKER_01: 24:52

Great to be here. I'm stoked. Legend. Absolutely. Beautiful animal hanging out with you all. I mean, it's such a such a reprieve. I mean, you should see my calendar. It's full of product meetings and sales conversations. So this is this is a great that's right.

SPEAKER_03: 25:07

You showed me your calendar last week, and that just made my stomach turn, man. This is a great respite.

SPEAKER_01: 25:12

I promise you that right right now. We're strategizing on uh continuing the conversation on how do we recreate our CRM for two different pipelines to manage leads from two different processes, and what are the sequences and what are the tags? So this is a this is a good respite. I'm excited to have some fun.

SPEAKER_02: 25:28

This is a nice break. Well, I'm gonna start off this this interview with a uh funny email. And I I I told them before I hit record that this is the first time uh I ever sent kind of a an an uh F you in an email. And uh I'm gonna share my screen here. Is this mine? Yeah, yeah, this is stuff between you and Carly, dude. It's a good response. So, look if if you're if you're just listening, just pull over and and pull out YouTube.

SPEAKER_04: 25:56

Oh yeah.

SPEAKER_02: 25:57

Oh god, this is great.

SPEAKER_00: 25:58

We have we have the documentation.

SPEAKER_02: 26:01

Oh look, there are two for 2021. 20 August 2021. Steven says, What in the hell it costs 25,000 worth of shit traffic? Allie, Charlie, pop on ATRS or SEMRush and see just how highly the pages from April and May are ranking plus traffic they're generating, etc. etc. etc. I basically said, look, not all traffic is good traffic. We're we're basically trimming the fat on pages. So this is for elevate holistics, which will kind of set the tone for like you know, the company we used to work together on. I basically said we're trimming the fat. Connor says, Yeah, I'm okay with the loss. It looks like it was a bunch of marketing terms in there, man.

SPEAKER_03: 26:36

Just lower hanging fruit, honing in, dialing in. Exactly.

SPEAKER_02: 26:41

Such a savvy marketer. So like I was like, oh shit, like the client, like, oh shit, like Steven's pissed. All right. So I included some screen grabs. Here's some keywords ranking. Okay, uptick there. Okay, that's great. These are the the uh um qualifying condition pages. Steven then says, This is why I fuck with you. I come at you and you come back with fire. And then I said, No, dude. Just sent the middle finger. And then you said, Oh, yes, bro, we're gonna make great great business partners one day.

SPEAKER_01: 27:11

Wow, good. Uh I always lead with I lead with um professionality, okay? Formality, uh being a professional is definitely what I what I lead with.

SPEAKER_02: 27:22

So keep in mind that was also in 2021.

SPEAKER_00: 27:24

So you know, you're acting like that was so long ago.

SPEAKER_01: 27:27

It was like it feels like it was a different lifetime, you know what I mean?

SPEAKER_02: 27:30

I kind of I mean, yeah, it feels like a different lifetime.

SPEAKER_01: 27:33

It feels like that's fair.

SPEAKER_00: 27:34

That's fair. It does feel that way. It does.

SPEAKER_01: 27:36

There's been so much happened, it's unbelievable. Now it's good.

SPEAKER_02: 27:41

What are you doing? What what is synth? Give us the lowdown.

SPEAKER_01: 27:44

Wow, that's so good. I used to I used to just say whatever I wanted in emails to people.

SPEAKER_04: 27:50

I don't do that anymore.

SPEAKER_01: 27:52

I don't do that anymore. I used to hold my vendors high, man, about high expectations, like show up or shut the fuck up. All right. And this just didn't wait.

SPEAKER_03: 28:01

You still don't, you don't, so you're not really doing that now? Are you kind of like, is it more polished and professional? So if someone's like underperforming, is it like a very polite email where it's like, hey guys, I think we need to reframe our expectations.

SPEAKER_01: 28:14

Yeah, saying things like we need to get to we get to reframe it's like, yes.

SPEAKER_00: 28:19

So that's some growth. Okay, that's some growth. Good for you, Steven.

SPEAKER_01: 28:24

Yeah, thanks. Thanks. Yeah, I appreciate it. Uh, because I always felt like the business environment is so stuffy and boring and formal. Like, why can't we just communicate like we normally communicate? Yeah. Um, and then I realized that I probably should be a little more polished because my wife doesn't appreciate if I was like if my whole eyes like, ah, babe, you said you were gonna get this Sunday, but you didn't. I think we need to reframe expectations.

unknown: 28:46

Okay.

SPEAKER_01: 28:47

She's like, she's like, okay, and he's dead, and then I'd be dead and y'all would be coming to my funeral. Um so with with Synth, I just took it over in July. It was a it's a it's a it's an AI company that helps automate um psychologist reports. So when I took over in July, it's coming out of what's called a Venture Studio. Um y'all are all probably familiar with the model, but for people who aren't, they have, you know, a venture fund has a small team, an operations team that gets fielded opportunities and they choose what they want to invest in. A Venture Studio has a team of researchers and a marketer, an engineer, a product person, you know, and the researchers basically go out and find opportunities and then create businesses around them. So they'll find design partners and start to deploy capital and resources around an opportunity and treat it just like a sales funnel where they're but it's more of an opportunity funnel. Um, and then as the opportunities get more real, then they'll start to incorporate this design partner, they'll sit in the office with them. And then once the business basically has the first MVP built, they'll bring on a CEO. The CEO then gets a chunk of the company and then goes and raises capital around the business. Once it's capitalized, it's basically spun out. of the venture studio and it's on its own, right? So it's actually a really tough model um to to operate because you have overheads of extremely the overhead of extremely talented team could be 10, 15, 18 people. So you have to you have to raise all the money to be able to to live long enough to then see these opportunities all the way through. So Synth, um it came out of a venture studio called 19 Days here in Tulsa. It's the third company, really the fourth the first one failed. The second one is just getting its legs. The third one's about to raise a seed. So really we're the fourth um company that's created coming out of it. They started doing all the research last November months ago the product hit the streets in May and then I took over in July. We had a few thousand dollars in revenue and I took over in just a few customers. Now we're I think this month we'll hit 30,000 in monthly revenue. So we're really excited we went from losing 70K a month to now we're we just lost 20K and this month I've added a few people so I'll probably lose about the same. So our burn is extremely manageable and we're growing at about 20% a month. So if I keep doing that I'm gonna have your break evening business. Your break even's like what the new year Steven I mean we'll know in February pretty sure yeah yeah and that's assuming that I don't cut some people if they don't produce you know 5x what I'm paying them then they better be producing that 5x value or they're gonna be gone. So or we don't bring on more people but I'm I'm saying I think that we're going to break even in January. Although I did just bring on Dell shout out to Dell compliance software real quick because those guys are managing everything. They're cheaper than what I was doing and I'm gonna have SOC2, GDPR, all the things so much faster. So um really excited about what's what's possible with this new technology not just Dell but AI in general AI native companies companies that are relying on agentic processes it's so exciting right for it's just changing everything. I can't wait to see the software industrial complex take a massive shit. And Dell is going to do what Steven that's like going to help you with compliance in regards to DPR, soft two, ISO ISO 27001 yeah they bundle it all they include the services it's really it's we're just kicking off so I'm really excited I might just be too excited and we'll see what actually happens and the rubber meets the road but compared to the others like Drata and Vanta it looks way better.

SPEAKER_00: 32:27

That's also branding all over San Francisco because I do live in San Francisco. But I am curious Steven because it's not just about like GDPR and SOC too but it's also the fact that um like there's a health sort of component and so like how do you deal with things like HIPAA? Like what is that like?

SPEAKER_01: 32:47

It's a part of it. So I mean they're going to it is okay they're they're managing that that whole validation the problem with HIPAA is it's a regulation it's not like a certificate. So you kind of make your case for what you qualify as and what you're obligated to follow under the rule because of are you a business associate? Are you a covered entity? We're a business associate of the covered entities and so we have to follow the breach notification law and the security law but not the privacy law but maybe the privacy law. So those things we have to talk to our attorneys regularly and get feedback on. Everyone thinks they're HIPAA compliant just by like signing a BAA and doing some encryption and it's just not true. So that's one of the things that actually keeps me up at night is is this compliance because like I don't know if you all just saw that that crazy Claude announcement that they used that they were using Claude to execute a cyber attack and Anthropic actually stopped it. Like that shit is crazy to me because like we have we have it's not it's not just like oh Connor got a cold and got some Amoxicillin like our the data that we have sensitive man. Really sensitive everything about their mental state and status and all the fucked up things that have happened to them in these reports. So this stuff really keeps me up at night. So I'm excited to have what I think is going to be a good partner to really speed this up and we can actually go international to we're going in her name, baby. That's awesome.

SPEAKER_03: 34:05

So Steven just out of curiosity too I mean I would assume you kind of had some of the knowledge and wherewithal about the HIPAA compliance stuff just based, I mean your your whole career really is is health right in that health industry telemedicine with elevate holistics you've done some stuff at craft as well so were you able to kind of bring a lot of that over into synth and kind of apply it or or was it more of a clean slate and you really needed to like you need to do something entirely different because it's just a different sector like well I'd say more clean slate because every other opportunity was outside of the insurance realm.

SPEAKER_01: 34:38

So really companies don't have to be HIPAA compliant if they don't do take insurance because they're not considered a covered entity. So like cash-based clinics don't really have to everyone just does it because that's kind of the standard for compliance and and it helps mitigate risk. So you know with the other companies everything was cash we didn't do insurance so we didn't have to follow everything we had policies in place but it wasn't like we needed to really worry about it entirely. This one is the exact opposite end of spectrum like we absolutely must follow everything to the letter of the law. And so it's been a big learning experience especially as I'm learning about well where is all these where are these policies come from it's like oh that's 10 grand a lawyer is here got to spend a bunch of time writing policies here. So I didn't have any of that existing elsewhere. So it was all it's all net new and it's been it's been daunting as like a non-technical non-security founder right kind of put this stuff together right just working with lawyers and other people say do you you don't really know someone's like yeah that's gonna be 12 grand it's like I don't know how do I know if that's a good deal or not right there's just so much information asymmetry with all these conversations. So um that's been a brand new brand new process and I've spent a lot of time thinking about it.

SPEAKER_03: 35:53

And and like when you made this you know when you made the switch over to to join with Synth, did you kind of know that going into it that that was going to be one of like the the main objectives for you to to check off the list so to speak?

SPEAKER_01: 36:05

I think it's gotten more and more important as I've been in it, especially seeing how much interest we have internationally and internationally security's taken even more seriously than here in some respects and you know you look at some of the frameworks like GDPR like it's just such a massive ball lake and it makes sense why it's harder to do business in these places when you have to spend so much money on compliance just to get really started even in the market. But that's one of the reasons I want to do it right and we have we have interest we have New Zealand Australia South Africa the UK I mean it's amazing to see so we know it has to be done so that we can open up those markets. Yeah that makes a ton of sense. Plus if we want to go enterprise right you kind of have to be SOC two type to even have an enterprise conversation.

SPEAKER_03: 36:51

So well we were kind of having this conversation you know you and I I know we spoke last week too about the the approach whether it's individuals or you go the enterprise solutions way and I mean you have to go that way right if you want to scale and you want to scale faster and make it worthwhile. So that makes a ton of sense and I I guess another question I have too is is what really drew you to to go along with synth right like was there some like pitch they gave you was it just kind of taking a look at and you're like yep this looks awesome I want to jump on it was there was it the AI piece that was missing you know what really convinced you like like were you always wanting to get in the AI space and that was it or is it just like I like the company like their mission and where they'd been and let's go.

SPEAKER_01: 37:31

Yeah a few different things we moved to Tulsa last December right to take over Kraft at a concierge medicine group um and wasn't the CEO I wanted to be CEO again and there aren't many software companies being built in Tulsa who would have thought you know it's like not exactly the innovation capital the world here you know unless you're in oil and gas. And so it was a combination of a few things it was really the only opportunity you know I took a a leader's like an AI leadership course machine learning course loved it absolutely loved the possibilities love what what what's just automation how do we have lean teams how do we how do we connect systems how do we use it to our advantage so that's always been really exciting um and I wanted to be in the space. So when I got and I'm invested in the venture studio so that when I I got an email in April that said we're hiring the CEO for this company I was like this me like this is me. This is me like I've never done a B2B firm before you know I've always done B2C stuff so Mallory I feel you but wanted to wanted to try my hand at B2B and it's and I've never had like just built a software company. So it was all a big challenge and I just continue to whoop myself I'm just like I'm just like it's like I like what can I do that's the hardest thing possible right now and let me just do that. I'm like oh why why do that um but you know it's never what was really exciting is that the lead velocity with the business is insane. Like we can all we're doing is advertising on Facebook and we're getting my calendars full nonstop I pay like three thousand dollars a month in Facebook ads and I'm just getting business. It's like it's it's the fucking craziest thing. So when I saw those numbers and I was going through the interview process, I was like, oh I was like first of all the sales process is much more like a B2C sale just because psychology practices tend to be one to three people. So it's much more like a B2B or B2C scale I'm sale I'm I'm talking to business owners, right? So I don't have the minutiae of trying to do big enterprise sales. I used to do that in the past I was no good at it. So from just a who am I talking to what's the velocity and the intensity of the need here I mean with when we're getting this many leads on this little of a budget, it's obvious that this is a problem. Like this is a serious issue that needs to be solved. And then I started then I got on board and started listening to these people we've had customers cry because they're like I can finally not spend my weekends like my cousin my husband's husband's gonna love this I don't have to spend my weekends. Yeah I've been on calls with three people that have teared up like holy moly like I finally can get my life back. Right one woman said like I I got a gym membership but I haven't had one in 10 years and I'm actually using it because I don't have to spend my nights right it's like it's all these awesome things feedback that you can get to really valid yeah everything that you're doing.

SPEAKER_03: 40:24

So Steven like when you joined synth too I mean I'm just really curious like the whole build was already there for then the reporting aspect of it it sounds like right the MVP was here.

SPEAKER_01: 40:33

I mean we were basically a user experience on top of some foundational models using some API calls right it wasn't like it was yeah I mean how much of how much of these AI businesses are really just that yeah um now we've single now we've changed it so you know that was my question how have you how have you kind of taken it and and transformed it since you've been in there like like have you made a ton of changes to that? Yeah for sure right we were doing all of our document analysis was basically Gemini and and then that built was Claude right so we were running API calls we were taking we were consuming the documents packaging it with our prompts that was the PDF was ran for data extraction on Gemini Pro. And then we ran all that extraction data plus our prompts into Claude Sonnet at that point it was three or three point five or something to to uh to generate the narrative and then the narrative spits out on the page and then they can use the AI to change different change the tables and change their whatever. So now um we use uh we basically moved everything into bedrock so that we can control and eat more easily switch out models. We've built extraction pipelines by file type. So that's a that was a really big change. So it detects if it's handwritten, send it one way, if it's a scan document, send it another way. If it's a doc X, extract it a different way. So those pipelines have been have been really cool and then we've to continue to try to we've continued to focus on more specialized models. So if it's a handwritten document use this model through this pipeline. So having that specialty and we've also um incorporated more open source. So you know right now the problem with hugging face is that it's pretty expensive. But like Quinn has by far the best OCR but it's the most expensive. So we're still trying to figure out how we're going to use this right so we've gone from just a couple models on the MVP to like six different pipelines models extremely specialized and just continue to add compliance features and functionality to make it easier for the users.

SPEAKER_03: 42:32

That's awesome. I mean yeah a lot of work so really a big transformation from when you first started to now and I'm and like you're saying you're gonna keep going through those iterations as well do you what's the feedback? I mean it sounds like it there's a ton of positive feedback here right from psychologists you sign up with it everybody's relatively happy with the report output or you have you gotten any feedback along the way that's helped you kind of I guess optimize the best word I can use like what that output looks like.

SPEAKER_00: 42:59

It's like it's it's the it's the optimizing of the output but it's also like how do you foster that trust with both practitioners and patients right like what is that like too?

SPEAKER_01: 43:10

Well yeah we don't interact we don't interact with patients at all. So just that's just just a practitioner yes the relationship is you know you're going through your process with your therapist or your psychologist and then as the psychologist and they're getting documents they're uploading getting the scent okay and then giving you the output normally there's some sort of consent that you they're using AI. The the trusting with with the data output is it's really tough, right? Because everyone expects AI to hallucinate a little bit but what's the degree at which it's acceptable and you know there's a disclaimer that they have to review everything but you only get a couple tries and it's kind of like debits and credits. You know, once that account is is at negative they're not going to use the system anymore. And then it's like pulling teeth to get them in right so as we've as we've made changes we've shot ourselves in the foot sometimes because we're trying to run as fast as we can only have one developer I've got ahead of psychology and me where we don't have adequate testing. We're not writing test cases right so we've kind of blown up some shit uh as we're making these changes and we've definitely lost some trust but the way that we get that back is just by I mean I'm spending a lot of time with our customers. I mean I am I'm giving credits anytime our anytime that we mess up, I'm giving them credits on their account because we we're usage based, right? So another cool thing about AI is that we're usage based and not just subscription. So we can easily give them credits back, but we're also our average customer is worth a lot. So our wallet share for these practices is quite high relative to software spend because they're able to like literally not have labor, which is like people are getting rid of transcription services and dictation services and ghostwriters and and contractors to use Scent and when they're depending on us it's great for the business model but it means we have to get our shit right and so when we fuck it up then we're very apologetic and well you know it's funny it's like your damage control I mean Steven damage control a lot. I get it I get it yeah a hundred percent and uh Steven you had mentioned you're you you guys are running Facebook ads and you know you you've worked with uh uh marketing agencies myself Connor in the past what have you taken from that in the past like what you've learned from working with agencies and marketing into what you're doing now with with synth uh what have I taken from the past I'd say right now like I've never seen paid ads work before more I've never seen really never seen it we spent 125000 on ads at craft medical Connor remembers none of them work we've spent so much money is that like a product market fit thing or is that just like uh shitty websites shitty flow I mean all the things or I mean Mallory would know like bad messaging perhaps on these ads no I'm just kidding it was it was literally it was a bad experience you think about it was just friction the whole way through right so right now one of the things we focus on is how do we reduce friction right like my whole as possible I'm I'm thinking about I've got it up on my board right now like are we building features that delight are we building features that are we reducing friction through every step are we anticipating their needs right like these are what we're focused on now. We didn't have that perspective back then so are we you know I mean I got that from one of you know Linny's podcast thinking through like if friction it's like how do I increase the velocity of the user getting into the system well it's the user experience but from these Facebook ads directly I am working with the art director the creative director is so she's fucking perfect. And like that is probably the biggest difference that I think other than like there's actually a big need for what we solve when people are searching for it. I have psychologists at home on the weekends just doom scrolling on Facebook and Instagram and then they see our ads like there's certainly a mix of all of that but I haven't actually worked at the creative director where I'm like wow that feels good.

SPEAKER_00: 47:09

Like oh wow what is that like what is that messaging like from a high level without getting into what exactly it is on the ads itself like is it are you speaking to the problem? Is it a problem versus solution type messaging? Like I'm I'm just curious like what's really working there.

SPEAKER_01: 47:24

Yeah there's there I think there's four ad sets running right now and it there's it varies. There's some like one that we some of the pain points like we know that psychologists hate doing data entry into tables. So one's about no longer doing data entry into tables one's about um report you know write reports in minutes not hours. You know one's about like regaining freedom right like there's so there's some of that but the best return right now I think is the one that's basically write reports in minutes not hours.

SPEAKER_00: 47:52

That makes sense. I um so my husband is a chiropractor and the amount of time he spends writing patient notes is insane. And in fact we were just having this conversation recently I'm like you could save so much time and get your life back if you didn't have to do these freaking patient notes. So I'm not in that field but and I don't understand but I'm seeing it firsthand in my life and I'm like you gotta there's got to be a way to solve for this now. So I can see why that works.

SPEAKER_01: 48:22

You I will I will tell you that one that would do him great to look at which is who we're kind of modeling our security as a strategy off of is is Heidi. If he hasn't seen Heidi Heidi's a great solution.

SPEAKER_00: 48:34

So Melly's on it right now she's like heck yeah she's like he literally I mean low key kind of had an argument like kind of had an argument about it because I'm just like there's so many tools out there now like you're just like he's like well I'm worried about it being HIPAA compliant I'm worried about this and I'm like yeah but you just got to do the research we got to like test it out. Like if I'm going to my vet and I'm going to like all these other places and they're already figuring out ways to incorporate it I'm like you as a small business owner like there's got to be a way to do it. Um so this is the other morning the conversation I needed to have today.

SPEAKER_01: 49:05

Well that and MD Hub so I'd say you could start you could start at those two and they're uh MD Hub's out of SF. I don't know where Heidi's out of but Heidi is the most compliant solution in the world. They've got a contract with the National Health Service of anyone you know they're compliant as well. So that's what we're coming after. Yeah I mean we're we're basically trying to do the same you know like we're we're rolling out we just rolled out dictation we're rolling out an ambient listening solution so we're doing what everybody else is doing right we find our wedge our wedge is the toughest pain point right now and then we're gonna build um horizontal on the workflow all right ambient listening AI agents right all those things.

SPEAKER_03: 49:44

Steam would you say like MD Hub and and then the other one that you mentioned is it high what is it Heidi Heidel? Heidi Heidi Heidi helps would you view those as competitors then to synth or is it a little different different market?

SPEAKER_01: 49:56

Yeah really good question. So I say right now because they help with notes but not reports they're really nothing like people are using them some ways I would consider them more of a substitute than I would competitor. I think about the large language models as substitutes as well they're like quasi competitors and substitutes because they're a tough they're a tough thing to sell against for someone who's savvy right you get a lot of people who just want the solution that's going to be the fastest most convenient right even though they're paying 10x more percent than they would ChatGPT you know because we're compliant and all the things um but I see Gemini and ChatGPT and all those as competitors but also substitutes and it's to keep up with them is tough like our user experience has to be top notch right as opposed to um because they're getting better and better all the time and those are the expectations that users have right all of our psychologists almost all of them use Chat GPT. When I come across a customer who hasn't used any AI yet I don't even know if I like it. I don't even know if it's a good thing because pumping is such a foreign concept probably not yeah at that point for you it's just such a foreign concept so it's um it's interesting right because someone's like I can do most of this in Gemini and it's like yeah if you want to build your custom gems and sign the BA and do all that's what it is then like totally yeah I should do that.

SPEAKER_02: 51:13

I I've heard a podcast I can't remember the podcast but I look with with with our AI app coming out you know there's there's APIs there's GA4 Search Console Google ads and it aggregates everything together and then you can you know do your your um uh data reporting that way and then a friend asked me like why don't you just upload to GPT and I referred to a podcast and I can't remember what it was but it was something like it's not purpose built. It's not like a direct fit for what it is that that person is trying to solve. Therefore it's more friction therefore that customer will probably just you know get distracted and do something else. And I think purpose built um uh solutions with a good brand that we've learned from Mallory and good paid marketing from you know Connor because I'm so deep in the SEO space like that's that's what I've learned has like scaled like like crazy.

SPEAKER_01: 52:04

The day that Chat GPT and Gemini and Claude and then actually raise their prices on consumers is going to be a day that I'm jumping for joy. Yeah jumping for joy.

SPEAKER_03: 52:13

That'll benefit you guys heavily. I think too though it's like you mentioned you know you could have any psychologist go on to a Gemini or chat or or whatever they want to use but you'd have to be really sophisticated in the space to know the prompts right to even put in there to get the output you're looking for. Otherwise I just don't think it's it's refined enough. Like I'm assuming again the way you've set it up is is you know you have some set of prompts or something that you you're telling it to look for specifically so that the output is unique one unique for each case and two actually providing valid information as opposed to I mean listen a lot of AI stuff output to me is junk when I go on chat GBT or Claude right you really have to validate it. It's slop it's slop is the best way to put it so definitely a good point there. I just think you'd have to be really really sophisticated in the space to even think about doing that as opposed to using something like synth but just my take.

SPEAKER_01: 53:05

Yeah it's true. I mean we have customers setting up like cloud projects but it's few and far between right the people that are really leveraging them well are doing just fine and saving a bunch of a bunch of time and money too right that's why but there's the compliance issues.

SPEAKER_00: 53:23

Info with chat GPT even though I won't lie like I sure a lot of people put in very personal information on chat GPT. So I guess that you know which is is it is silly and it is scary actually um how much personal data people probably place into chat GPT.

SPEAKER_01: 53:40

So I wouldn't trust my patient notes um to be on there what if that got leaked exactly and you know too yes if we want to search right we should probably all go on chat GPT and be like tell me about Steven Steerman.

SPEAKER_00: 53:54

Like that's a bit about the emails no because like literally two years ago if I looked myself up there was nothing and then I look myself up and now it has all this information about me. I'm like where did it get this?

SPEAKER_01: 54:07

Is it LinkedIn like is it something weird so yeah I'll tell you what I'm thinking a lot about what I'm thinking a lot about right now as we're building our first self-service pipeline automations is like how do I wow my customers in the first like 60 seconds right like how do I get them to feel the magic in 60 seconds you know where like synth is amazing because people most of the solutions none of the solutions you can actually upload all of your documentation right we have people that upload hundreds of pages of stuff and then you have a report right and that was the really great thing about the early but as we're as we're building it's becoming more convoluted right like where do we go how do I build my templates how do we do this because people are asking for it but like maybe just the 20% that are the hardest to please another 80% don't find it's even necessary. So like how do we actually build a solution you know gamma's CEO talks a lot about the company was basically going under and then they spent two quarters optimizing for that first 30 seconds in the product but we have something different right people have to give us documents or whatever well how could we build this where maybe there's just like already loaded sample docs and then they get to watch it generate a report right like how do we that's cool that first 60 seconds like welcome to the account like something report. Yeah right that's the ticket.

SPEAKER_00: 55:25

I think that consumers and just people in general are expecting their first time experience with any AI tool to have that magic moment, especially with the rise of ChatGPT and Claude. And so I could see them just kind of being like if it's not there then like I'm churning um even though there's a lot of value in what you're building but I think there's a lot of expectation around having that magic moment for sure. I I mean my time at Bolt was that right like you're seeing the code you know show up in real time. You're like what is this magic that's happening in front of my eyes as yeah yeah crazy. Yeah.

SPEAKER_01: 55:58

Just I mean the the intuitive every time like Chat T B5 and then the agent mode came out and like I spent a lot of time in these things and a couple months back when I was in agent mode I was like oh my God this is amazing. It's like would you like for me to do this task? Would you like me to perform this task next this seems like the way you would like to go is this true I'm like yes yes yes yes yes give me that and how do we that's what I mean like people have to have have to experience this right especially this is these companies are setting the bar right and with our experience especially ChatGPT is trying to force people more to a a voice form right like engage with the voice do we need to do that too right like how do we yeah like what are what are these new form factors coming out? How do we how do we you know do we what's the most forward looking state start there and then build the technology backwards like computer use is going to be here in a big way in what nine months maybe like Claude's Claude was doing basic computer use like 12 months ago like this has to be about to be released at some point and then can people just talk to Synth like hey I'd like to start a new report for Steven Steerman birth date you know gender is this go into my files pull in everything from this folder and then bam and then you can just voice and tell us what's going to change like that's where I really want to go but I wonder if users are even like ready for that right so is that like your your two year outlook for synth is really like again you're saying working backwards.

SPEAKER_03: 57:22

So two years from now that capability is most likely here already and and how you leverage that for synth moving forward.

SPEAKER_01: 57:28

Yeah totally and I'm like why do I even need to be able to do that you know you think about this experience with doctor's offices I was just in the doctor's office today and they still have paper forms paper fucking forms.

SPEAKER_03: 57:41

I'm like this is crazy you could have still 2025 guys you know they're faxing stuff probably they're faxing I had a dentist office ask me to fax something at like my insurance chart and I'm like I'm like what I'm like I don't even I don't know where to go for a fax machine. I'm like I'll just I'll just drop it off guys like is that better like it was nuts. Couldn't believe it.

SPEAKER_01: 58:02

So we you know we think about you think about what an intake process can look like today like it should be a voice agent that's dynamic with decision trees that's saying is this what you're experiencing and it's saying well that's not it that's not it let's put in some screeners let's you know so that's really where like as we want to create more of this workflow it's much more of an agentic workflow clinical sidekick with a psychologist right where it's like all right Mallory is a new patient and we just basically press go and the voice agent engages starts collecting information from you via text from a call it's going to produce a much faster experience because there is a user experience problem on both the patient side and the clinician side when it takes eight weeks can you imagine if it took eight weeks for you to get an MRI result you'd be like this is crazy what's going on here for sure.

SPEAKER_00: 58:47

And then it's like$200 a session at least right so like how much money did you just pour into a session every single week right before you even get to that.

SPEAKER_01: 58:55

Yeah. Give me this feedback right like I have an education what if I have an education accommodation that I need and my psychologist is taking eight weeks because they've got a bunch of new stuff and report writing takes so long. So we could just solve all of this over time.

SPEAKER_00: 59:10

It's not just about the magic moment, but it also seems like it's it's about how do we take these practitioners that are used to using technology from the stone age and educate them and give them a super easy way to onboard because we know once we do that like they'll be hooked right like they're gonna want to like of course who's everyone's gonna want to you know stop using fax machines and paper forms. So right so it's about that too I think you'd think I think well you'd think you're right you're right.

SPEAKER_01: 59:37

And where are we? You know how did bolt you know how did bolt maybe you were not even there yet but like through this you've got your early innovators right and then and then you've got your you know what is that then you get to the innovators dilemma then you get over to this early majority like how you think do you think bolt has bolt reached the chasm to the early majority yet I don't think so still no I think there was a stat that was shared.

SPEAKER_00: 59:59

This had Be a few months ago, but it was something like 10% of Earth's population has utilized Chat GPT. So if you take that as a stat, drill it down to all it is, it is, it is.

SPEAKER_03: 1:00:11

That's lower than I thought it would be, though.

SPEAKER_00: 1:00:13

But if you compare it to like how many people are consuming social media platforms and the rest of the internet, it is technically low. Oh, interesting. And I even saw another stat in that most people that are using AI are making over$100,000 a year. So the people that know how to use these tools are people that already come from a background of like having probably the privilege of being able to like get easy access to these tools or learning these tools or what have you. So, and then with the vibe coding space, I think a lot of that space was actually marketed towards like indie hackers, developers, like you know, folks that want to become founders. I talked a little bit about prosumers with you, Steven, before we even started recording, which to me is like, you know, someone who wants to become a content creator, they buy all the gear, but they've never actually been a content creator before. So they'll buy the fancy camera, they'll buy the like nice mic, right? They want to become an expert in that and they're starting to purchase things that like will lead in that direction, but they don't have the knowledge or the skill set and are seeking that out. Um, but I will say, I think that we're not even when it comes to that curve, like we are still really early. Um, because if you take that 10% stat and drill it down to all of these different tools, it's pretty small. Um, so I think that there's still, and then on top of that, too, I think like what I noticed, and this is probably not really something for you to think about, Stephen, but I think there's a lot of young people who are anti-AI. And so, like, how do we change the narratives and the sentiment around AI? I think there's a lot of work to do there as well before we get to that, like we cross that chasm.

SPEAKER_03: 1:01:46

Absolutely. I agree with you, Mallory, too. Again, it's it's how do you use it as uh, you know, an agent for good, so to speak, as opposed to just again kind of slop output for whatever it is that you want to find out or research. So wow.

SPEAKER_01: 1:01:58

So you you think that we're still in the because it's it's innovators and early adopters and early majority. So you do you feel like with AI in general, we're in the early adopters still and haven't really crossed that to the early majority? Maybe innovators.

SPEAKER_00: 1:02:12

I mean I don't know. That's a this is a great question. I'm I don't know, somewhere in between. It's somewhere in between there, you know. Cause like I don't know, I live in my own bubble because I'm in San Francisco. Every ad I see is around AI. When I hate it because I'm like any of this shit mean, like, I don't even know. I've done the billboards, yes, for bolt, but like what does half of this stuff mean? I don't even know. And I work in the space, so that is like it was it's whatever. But until like we're hitting like mainstream, I think that, and we're not there yet, I don't think, with AI. Maybe you could argue maybe Chat GPT has kind of like been the start of that cultural phenomenon, but I think there's a lot of work to still do there. Um we'll see.

SPEAKER_01: 1:02:56

I'm I've got my ex algorithm completely dialed in on the SF bubble because because I've just yeah, we were gonna raise capital until I was looking at all the firms, followed all the firms, follow the GPs, followed YC and a bunch of theirs. So like the algorithm was just and to see what these valuations it is crazy, right? They're drawing comparisons to the dot-com bubble, but they're like, there's not a single dark GPU in the world. They're like, how can that be like it's wild what's happening in the space and just see all the innovation just like Delve or just like some of these companies? I just saw the most amazing looking CRM that was just that just came out and uh which I'll I'll I'll find it. It looks so slick, and it's basically giving you the data when you need it and creating your profiles as you need it instead of this static interface.

SPEAKER_00: 1:03:46

Love that.

SPEAKER_01: 1:03:47

That's cool. Oh, it's so exciting.

SPEAKER_00: 1:03:49

I don't have to use HubSpot anymore. I would be like, Light field.

SPEAKER_03: 1:03:51

It was Lightfield, right? Lightfield said that to me uh last week and it looks um insane. Slick.

SPEAKER_01: 1:03:57

I mean Lightfield. That's what I mean. It's like just the possibilities for almost you know, if manufacturing had just in time manufacturing like 50 years ago, it's almost like we're just in time user experience is now possible. Like that's fucking sick. Like, get me that.

SPEAKER_02: 1:04:12

That's about us. Uh, Steven, as we wrap up here, whoo, we are cutting it tight. Um, you are not a I want to ask you one more question real quick for our audience. Our audience basically vibe coders and um builders. You're not a first-time founder, you're a successful CEO. What's one bit of advice that you could offer somebody who is currently building and would transition into a CEO of an AI company? What's what's one bit of advice you could offer them?

SPEAKER_01: 1:04:37

Somebody that's vibe coding and then to become like to start their own company? They want to scale it, yeah. Scale their own. Oh wow. I don't even know what it is. No, man. Yeah, no, we're totally advice. I'm still I'm still figuring this out myself, right?

SPEAKER_03: 1:04:51

Just work with the right marketing that's honest. That's honest.

SPEAKER_01: 1:04:54

I'm still figuring this out myself. I mean, it was a big, it's a it's been a really big transition to go from elevate to concierge medicine and now running this company where like we don't even have a product person. Um, and I think what I've learned a lot from my senior engineer is like really this is like design is really important and ultimately creates friction within the organization. Like we need to be putting out POCs earlier and more often. And then as we track the user data, then we make improvements. So, you know, what I was trying to do was all right, cool, let's come up with this thing, we'll get the designer involved, and then in a week we'll have designs and we'll give feedback and then he's like, screw that, let's put this out, track the data. People may hate it and we'll stack it versus putting time into this. And but I'm I'm I'm conflicted in that in a way because not that it's necessarily a moat, but like taste and design now and user experience is so important as a first impression. So, how do we balance like the need for speed and also the need for beauty? And I think that's been uh a big conflict that I'm uh dealing with right now. Um, also how do we uh how do we balance like new features and functions to get new customers versus uh going really deep and dialing in the current experience with limited resources, right? We're not funded, we've got one dev. So how do we balance that plus risk mitigation? Like we think about this compliance stuff, like it's gonna take 20% of our time. Well, that's 20% of the time we're not moving features and functions forward for new customers or the new new user experience. So balancing like growth, near-term revenue, like building the things that give me near-term revenue versus long-term revenue and trust in the market. I'm still figuring this out. I mean, I think about this all the time. I'll I wake up, I don't sleep as I'm thinking about this stuff, right? And that's like that balance is uh tough tough to really nail down for me.

SPEAKER_02: 1:06:48

Gotcha, gotcha. Well, um, Steven, thank you for taking your time. Uh, we appreciate it. And uh, where can people find synth and stephen?

SPEAKER_01: 1:06:57

Synth.ai. And then my email is ss at synth.ai.

SPEAKER_02: 1:07:02

So cool, cool. All right, and we'll be back next week with a new guest, Pete. Thanks, Steven. Appreciate it, man.

SPEAKER_00: 1:07:10

Thank you.

SPEAKER_02: 1:07:11

All right, and we're back. Uh, great interview with Steven, the beautiful animal, as as we called him.

SPEAKER_00: 1:07:17

He really is. I have now been able to understand what you meant by that, and I got it. I got it.

SPEAKER_02: 1:07:22

The uh uh golden retriever energy, great dude, sharp dude.

SPEAKER_03: 1:07:26

I love how that was he mentioned that too to us, literally said golden retriever, yeah, like 30 minutes after I had mentioned about his personality, which which he is. You know what though? It's great, dude. Awesome dude. It's really, it's really fun and kind of um I'm looking for the word. I don't know. It's just nice to really see growth though, right? With people that you've worked with for so long and colleagues and just kind of where he was when when we started working together and where he is now and the way his mind thinks and works. And it's just it's cool to see that transformation in people, right? As you work with them and as they kind of go down their own path. So super insightful conversation with him and some deep insight into what's going on with synth, which I thought was really cool.

SPEAKER_02: 1:08:06

So I uh I I shared that story with Kaz uh over the weekend. Kaz is our our old COO at the time, and uh I I showed her that email. I told her the story of the email, she was, oh my god, you you had you have to bring that up, you you have to share that email story. And I was like, Okay. So I I pulled up that you know old email thread.

SPEAKER_03: 1:08:23

Did Kaz and Steven ever like like converse much or not really?

SPEAKER_02: 1:08:27

Yeah. He goes, I love her. He goes, first of all, I love the accent, and I was like, uh it weighs on you after a while because she it's just a very irritating voice. Charlie. Yeah, choli, get your shit together, Charlie. Yeah, I heard that quite a bit. Uh but no, she she's great, but yeah, uh Steven loved Cassie's like, Oh my god, he said the word he used was Kurt. He goes, I love it, I absolutely love it. So uh great interview. Um, big things from that dude. He he's gonna scale that company, like like no question, he will scale that company.

SPEAKER_00: 1:09:00

Yeah, um, I mean, there's already signs of it, right? I think there's a definite need for it. Um, so I'm excited to see him build in that space. I'm excited for that space in general. And I've learned about a completely a few new tools that I sent to my husband as we were speaking to him. I'm like, you need to look into ID. You need to look into this. Like, if you don't, I'm gonna look into it.

SPEAKER_02: 1:09:20

Yeah, yeah. Well, good stuff. Uh, okay. So we're gonna start this week and we're gonna uh given you know our our efforts in the AI space and building and and marketing and you know, Mallory's past expertise and in the community and branding, uh, we're gonna start talking about different topics in the world of AI. So we'll just say like this week in AI is the again a working title. Um and that leads to our our rebrand, the AI Code Central, which uh uh was created by an SEO guy. So that's okay.

SPEAKER_00: 1:09:53

You've done great as an SEO guy.

SPEAKER_02: 1:09:55

A couple episodes ago, uh, we went through the rebrand process. You know, Mallory pulls a curtain back on what goes into that effort. So, Mallory, what what can we update with the rebrand efforts?

SPEAKER_00: 1:10:05

Yes. So right now we're in a place where we need to kind of create a short list of two to three names that the three of us are really excited by. And then from there, what I'd like to do is work with um an amazing graphic designer uh that I've loved working with named Fiona to kind of create some sort of visual directions for us to go down. Um, and then from there we start actually working on updating everything. Um and so yeah, this is where we're at with names. I think last week, Charlie, you mentioned that you were really interested in the name Launchpad. Um, so that is uh something that I was kind of playing off of. Okay, wait one second. Okay, okay. So launch pad AI was, I think we talked about launch pad. And then what I did was I kind of just like added the AI twist to it. I've been kind of back and forth on do we use AI, do we not? Um A, because like it's of course it's an important keyword, but it will differentiate us. But at the same time, I'm just like, I hate putting it in the name, but I'm open to it. Um, and so this first one, yeah, Launchpad AI. Um, I don't think that there's any other sort of mainstream podcast that um covers this name. What I like about it is that to me, it covers kind of the fact that we're for builders that are looking to scale um their product, right? Um, and do that marketing. So to me, launch pad kind of insinuates that. So that's the first one. Um, and then you'll see that some of the other names have to do with the word build. So the next one is just builders foundry. I think like the word foundry itself talks about like bringing new products and skills together, right? Like this is the age where people are building brand new products. I mean, Charlie, you're in the midst in the process of building campaign pilot, right? Um, and so uh, although it's not like the sexiest word, I would say, um, there's something to that. And then um, the next one is AI brand lab, which I actually want to do AI Marketing Lab, but I from what I was looking at, it seems like there's another podcast that may already exist that's called that. Um so, but brand does insinuate marketing focus. It to me, it's not just about the design, it's everything that we've been talking about since I've joined this podcast. Um, so I think that could be an interesting play. And I think it's not that far off from AI Code Central. Like it's still like, you know what I mean? It still has like three, three words there. Um, I think it's just kind of more yeah, high level is what I guess I would say. Um the next one is called Builder Sandbox. Um, so it kind of captures like the playfulness, I think, while at the same time, like to me, a sandbox insinuates like experimentation, right? Trying to learn new things and having the space to do that. Um, and to me, that's kind of what we're doing too. Like we are marketers that have expertise in the space, which is great, but we're in this entire new landscape. And so um, we are sort of taking the playbooks that we've written and we're kind of flipping them on its head and we're learning new things um as marketers, but of course, as uh um, so do our users, right? And then some other ones. This one's called brand launch crew, kind of similar to there was another name called Build Crew that I was playing with. Um but yeah, I think this one's kind of self-explanatory. It's about marketing, it's about launching things, and we're a crew. So um, not my favorite, but put it on here, put it on here so we can react to things. And then the last one is called AI Marketing Makers, which is definitely a mouthful, but to me, it it like perfectly kind of covers what we're doing. Um, we're creating a podcast in the AI space. We are all marketers, and to some extent, we're all makers, right? We're building, we have our own agencies, we're building our own products, we're talking to or we're interviewing founders who are makers themselves. Um, so that being said, that's kind of the list that we have here and where I wanted to at least start from and hear what you guys think. You could hate all of them too. I'm that's fine. No, this is just for this is to get us to get to like a level set place where it could be like, okay, we like this word, we like that word, like how do we take the feedback and then yeah, shortlist it to two or three, you know?

SPEAKER_03: 1:14:22

Always so big on just like the how things are like enunciation and just how smooth it comes out, right? Like you have a great point. The bottom one, which is what that was marketing, AI marketing makers, it's a little bit of a not too much so. I think the one thing though, too, and and you guys tell me, right? It's like I'm always in marketing mode, marketer mode mindset. I don't know if like builder, sandbox, I love it. I absolutely love that one for some reason, but it's like it it doesn't encapsulate the marketing aspect for me, right? It's more about the builds, and that's kind of where it would start and stop. I wouldn't get that just from reading the name of it, but I don't I may just be thinking way too marketing forward and not you know not about anything else.

SPEAKER_02: 1:15:03

Which so uh of these of these names, Connor, because I I've been staring like like I said, I've been staring at this for months. Uh, which one stood out to you first?

SPEAKER_03: 1:15:11

Launchpad AI, Builder Sandbox, um, builders foundry somewhat, and then the last one just because it has marketing in it, so it's like those four, but if I had like, and then I can order them if you really want to know, but no, that's helpful. Yeah, I mean, uh those are the four that stick out. Launchpad AI sounds great. I just you know what for me for some reason, it's like I look at Launchpad, and again, I'm doing all this go high level work, and it's like they have something called Launchpad within it, and I'm just immediately like go high level launch pad. It's like one of their selections in the CRM to start on the launch pad page. What does it do? Oh it's just like it's like the your home screen when you log into the CRM. They call it their launch pad. So but again, I've been working in that for the past two weeks. I don't think everyone's gonna think that. It's just the marketing aspect for me is is just missing in Builder's Sandbox, but I just love the playfulness of it, right? It's it's a little less serious.

SPEAKER_00: 1:16:05

Yeah, it is. It's one of my favorites too. But you're right, it is missing the marketing aspect to it that something like a launch pad AI has. There's no AI play here either. And I can even just I could play with this if that is like a direction we're excited by.

SPEAKER_03: 1:16:19

Yeah, I mean, I love that one. Builder sandbox, really. Again, it's like how do you just merge something in that either that captures the AI and the marketing piece of it? I think the AI probably being the more important. Well, builders kind of does speak to AI though, no? Somewhat? Maybe not.

SPEAKER_00: 1:16:34

Yes and no. Yes, yes and no. Yes and no. I think in the space that we're in, though, it's insinuated or assumed that all builders at this point will be using AI um to some extent, especially when it comes to building. Um it doesn't have to be vibe coding, right? It could just like there's so many ways to use AI to build.

SPEAKER_02: 1:16:52

So um the one that stuck out to me the most. Can you scroll up when you're done there? The one that stuck out to me was Launchpad. I like that one. Launchpad AI. Because that also if if we can tie in marketing there somehow. Um, and there's there's no mainstream podcast with this name. I think that also ties in our um uh investing arm that we're gonna be doing as well. So launch pad, uh like a takeoff kind of thing. I I don't know. We can play with that how we want. Um I think that um uh that ties that in.

SPEAKER_00: 1:17:26

Okay, cool. Cool.

SPEAKER_02: 1:17:27

I like launchpad, yeah. That would be okay. Um and then yeah, I mean, I guess after that, after I guess after this, we have to find domains. Actually, you know what? I'm gonna make note of that. You gotta find domains and then social arms and um and all that, make sure that's all clean.

SPEAKER_00: 1:17:47

Okay, so I think what I'm gonna play with then is like launchpad AI and Builder Sandbox. And I feel like those are the two to play with, yeah, and have iterations of those that kind of address the stuff that you guys are talking about. And then from there, I'd like to move into let's figure out how we bring this to life. Like we have a mood board, we have some values from our workshop. Like, I want to then take that and run in like three different directions. Sweet. Especially based off the name, because I feel like a name can also help it like uh give some direction to where you can actually take uh the brand itself. So that's why I kind of want to figure this out first because it'll kind of in my mind set the tone. Absolutely will.

SPEAKER_02: 1:18:29

Oh, I get it. Yeah, but like like Builder's Launchpad, oh yeah, that kind of I don't think so.

SPEAKER_00: 1:18:36

I don't know, I'd have to look in to see like what other podcasts like you know exist with this name. But um yeah, I think we can I we can find an iteration of this, I think. So okay, cool. Let's go. But yeah, that's the update. So next week I will be like, okay, these are let's like lock down a name in our next podcast episode because we'll have some uh new iterations from these two, and then let's lock it down and then let's move forward with figuring out everything else that's to come.

SPEAKER_02: 1:19:10

So things get real, okay. Is that when you sing the intro?

SPEAKER_00: 1:19:14

Yeah, that's why I'm gonna sing the intro. Now that I have this fancy new mic, I've um thank you, Charlie. Oh, yeah.

SPEAKER_03: 1:19:20

Thanks, Charlie, for the stand and all that good jazz. I'm not worried about this thing falling over now next to me.

SPEAKER_00: 1:19:26

No, it's great. You know what?

SPEAKER_02: 1:19:28

I actually don't have my stand. Yeah, we we can't sound like pigeons when we talk about this stuff. It's like that was the only thing going through my head. We've got to has have nice gear.

SPEAKER_00: 1:19:36

Yeah, I agree. Um, but I've been playing around with OBS and a tool called, I don't know, it's some sort of synth like reverb type of tool to figure out how I can.

SPEAKER_02: 1:19:49

Oh, you mute. Oh, she's muted. When she's talking about the synth tool.

SPEAKER_00: 1:19:54

I was like, Oh, sorry, I touched the thing and then I forgot. That's what makes it mute. But I spent like I didn't mention this earlier, but I spent like three or four hours uh during the weekend just testing out different mic settings so that I could sing specific songs, and then from there I will then figure out I'll make a little intro situation.

SPEAKER_02: 1:20:12

Okay. So we got a new name, we got a new intro, we got new colors, we got our pets' heads. We're getting there. Okay, cool. We're getting there. Let's do it. Okay.

SPEAKER_03: 1:20:23

Did you just say our pet's heads are falling off?

SPEAKER_02: 1:20:25

Yeah, it's from uh dumb and dumber. That's amazing.

SPEAKER_04: 1:20:30

One of the falling off one of the best movies.

SPEAKER_02: 1:20:33

They uh a little sidebar here. They played that at Red Rocks not too long ago, and it was just a it was so awesome. Oh, that's awesome. Red Rocks is so awesome. Anyway, okay, uh, so that's the rebrand efforts. All right, let's jump into um uh AI rappers. This is always going to be a topic. And now I've been the first project I made was a forecasted SEO uh uh tool that helped in our sales process in our agency because we would get asked all every single sales caller when are we gonna get the money back? We don't care about rankings, we we understand rankings is the mechanism, but the end result is more income, more more sales. When will we get our money back? Or at least 5x. At the very least, when are we gonna 5x? And we made a tool, I stumbled into it and I made this thing, and then I learned about rappers where we're basically on rented land. And that I I really did not like that. I that just sat uh really chapped my rear end, as my old sales manager would say. So um, Mallory, look with with with your you know experience at Bolt, um, what are your thoughts on rappers? And do you see do you see that as bad or good, or or like do you think there's a time and a place for rappers?

SPEAKER_00: 1:21:53

I mean use case, no use cases, MVPs, mock-ups, prototypes, early stage products, you know, like that you're trying to kind of test really quickly and get out into the market, I think it's great, right? Like that is what this technology has given us the ability to do, like like basically build and ship faster so you can get that data. But as like a long-term product, I don't think so. I think that at least within even just like the vibe coding space, it's such an example. When I first started working at Bolt, I was familiar with maybe like three, four to six, maybe max specific types of vibe coding tools. Lovable, replit, v0, base 44, windsurf, bolt. Like that's when I started there, which was in the spring of earlier this year. By the time I had left seven months later, which is not even that long. I mean, I was going on LinkedIn and there were new in our market watch channel and Slack, we're constantly pulling posting like new vibe coding tools that just came out of nowhere that people are building. Um, and so it what happens is that it becomes really difficult, I think, to create that moat and that like that differentiating factor that makes you different from not only all the other AI wrapper type of tools you're competing with, but then the chat GPTs of the world, the claws of the world, right, that are most likely going to kind of either build in that direction or support that direction. And I've I've heard a lot of people tell me, like, well, look, you know, like once you're for everything, you're nothing product wise, right? And I actually agree with agree with that to some extent. But like if consumers are already using Chat GPT to like build a plan or, you know, write a script, why would it be that crazy for it to then build a website for you or a mock-up of a product as a starting point? It to me, it's it wouldn't be all that crazy. Um, so I don't know if that argument kind of works here, but I do think that as a long-term sustainable business, building an AI wrapper is not it. I think you guys probably know this too. Like, I think any any business that's built or relied on like one or two platforms is going to be a death sentence. That was exactly my argument.

SPEAKER_03: 1:24:02

Yep, no, exactly what I'm thinking too. You're so dependent on those one or two platforms to make everything else function. That it again, it's not sustainable. And it's also, to me, it's not scalable in that sense either. Right? So you have two blockers there, but again, that's up for debate, I guess, but 100% agree.

SPEAKER_02: 1:24:20

I I think you should own that customer interaction from A to Z. Like all of that should be your land, you know, with with your castle on it, with your moat around it, and you shouldn't have rented land on there. So to me, um the the plan with campaign pilots is yeah, it's it's a wrapper, but I think as we scale and you know, the moat that Mallory just mentioned uh is your community, is your your brand, and um you know, paid ads, all you know, SEL, like that all goes into your your app. Yeah, um, real technology behind it, like native systems.

SPEAKER_00: 1:24:59

And then exact that was the missing piece, right? I was gonna about to finish your sentence was like all of those things are great differentiators, but then also on the product side, it's like you need to have that like one differentiating factor between you using a tool like you know, campaign pilots and chat GPT as an example.

SPEAKER_02: 1:25:17

Yeah, yep, 100%. Okay, gotcha. Uh okay, moving on. So let's jump into uh open AI spying on you. So this came from a short that I saw on YouTube of Jason Calicanis. Never met the guy. Obviously, he's a brilliant angel investor. Uh is he the biggest name in in Silicon Valley, Mallory, as far as angel investors?

SPEAKER_00: 1:25:40

Probably one of them. I mean, there's there's a lot of them though. There's like Peter Thiel, there's like uh there's a bunch of names. Um Reed Hoffman, there's you know, Mark Andreessen, there's yeah, Jason Kilconnet. Like there's there's yeah, quite a few of these folks, but I would say he's definitely like one of the more well-known um angel investors and well respected, I think, within the tech space for sure.

SPEAKER_02: 1:26:08

He was also a judge at the um uh your your build-a-thon.

SPEAKER_00: 1:26:13

Yeah, yeah, yeah. He was a judge for that too. Which is um is huge. Which is great. Yeah, yeah, yeah.

SPEAKER_02: 1:26:17

And his short is saying that open AI is basically spying on you, which of course, yeah, again, you're on rented land, you're using their API for like maybe a single shot API call or or whatever it is. Uh, they're gonna use to my understanding, uh they will use like the top use cases, I guess, or like like the the best use cases and just take it.

SPEAKER_03: 1:26:39

Totally rip it, I think is absolutely nuts.

SPEAKER_02: 1:26:42

Yeah, I mean look, uh look, if if if I put out campaign pilots and it was it was my native technology and like I wanted to build like an ecosystem or like an operating system, which is I I guess what they're building. I mean, yeah, I mean it's my right to do that. Why wouldn't I? Of course I will.

SPEAKER_03: 1:26:57

Yeah, yeah. I get it. I just think it's also right, like the you know, the capability for on like the open AI side, right? For to spy on something and spy on a build, it's like they have the resources and the means to get it done 20 times faster than like somebody else. And and that's where that's where the issue rises for me, right? From an ethics standpoint, is it's not really a fair shake, and you're monopol now, you're monopolizing the space too, right? And that that's where it really becomes a problem. But you know the behemoth, yeah.

SPEAKER_02: 1:27:28

Well, it's it's the poker at the it's the bully at the poker table, and like you just nudge everybody else out and uh just take all the chips.

SPEAKER_00: 1:27:35

So yeah, it's crazy too, because I think that uh and I'm I'm I feel like maybe I'm speaking for myself a little bit here, but and I'm not saying to do this, but I've put so much personal information in Chat GPT, like when it comes to my health, like very yeah, very specific things. And I like open AI has this access to data that like no other company would like I just feel like yeah, people are so willing to share everything that they want about ChatGPT so that they can get a deeper understanding of themselves, and like that to me is great, but it's a slippery slope of like this now being the problem.

SPEAKER_02: 1:28:11

Um so first thing that popped in my head is like, well, okay, I'll just build with anthropic then. But I'm sure totally like you just do the same thing, right? Yeah, or you're it's like you're you're just kind of hopping around with APIs. Uh plus I I'm I'm partially anthropic personally. Like I've I I think it's it's easier to build with, I think. So um, all that to say, don't be on rented land. I guess it's like paid ads in SEO, right? Just do that.

SPEAKER_03: 1:28:38

Somewhat. I mean, yeah, somewhat. I think a little, you know, you can kind of make the the connection there.

SPEAKER_00: 1:28:43

But I mean, because you're dependent upon these specific platforms and their rules and their guidelines and et cetera, et cetera. That is like a great example of that, especially when it comes to paid ads and how honestly, it's such a pain in the ass. Sorry, and not to rant about it, but I feel like you need to have like a frickin' paid PhD to just navigate the business manager with Facebook. But the the point what I'm trying to say, yeah, I've never you know what kind of brought that up.

SPEAKER_03: 1:29:09

I've never done it. It's well, they also not to I know we're gonna go, I could go off. I'm like, oh, it's ours because they just build, they just build on top of what's already been built instead of instead of going in, refining it, pulling out faulty, crappy code. Uh it's like it's just it's just continuously billing on top of it. And eventually there's gonna be a point in time where that they're gonna have to go in there and and clean all that up. I just don't, maybe they won't, I don't know. But it is a terrible, terrible platform in terms of how it's built out and how you have to navigate through it. I I've used it for 10 years and I can't even find certain items in it. It's crazy.

SPEAKER_00: 1:29:47

Don't get me started on this. It's it's uh anyone to anyone who's listening who has to dive into Facebook ads manager. I'm I'm sending so much, so much love and positivity your way because it's honestly terrible. So Um, but what I was gonna say though is that like you you like as a you know paid marketer, you are sort of dependent on these platforms to you know to some extent. So um I can see that that connection.

SPEAKER_02: 1:30:14

Okay, yeah, yeah. Uh moving okay, last topic moving along. Um A16Z, Andrees and Horowitz launches new media fellowship.

SPEAKER_00: 1:30:25

Yeah.

SPEAKER_02: 1:30:26

Mallory, this is your space. This is this is like you're the first one. Look at Mallory smiles. She's like, heck yeah.

SPEAKER_00: 1:30:33

I know. I mean, I'm smiling because like I thought about applying to this. I had some folks tell me about it. So I'm like, ooh, this is really interesting. Yeah.

SPEAKER_02: 1:30:42

Can you summarize this? And uh and what what are your thoughts on it?

SPEAKER_00: 1:30:45

Okay, so Andreessen Horowitz is launching a new media fellowship program. And it's an eight-week program for operators, creators, storytellers, marketers that are responsible for really sort of shaping the future of media. Um, and so as like if you're a part of their program, you'll get sort of insider access to sort of the A16Z ecosystem. You'll get to like, you know, be a part of exclusive workshops, you'll get to like learn alongside ambitious founders. Um and yeah, I mean, I think this is really interesting because um, and this to me, like this actually kind of draws back to to our conversation um with Steven, but like taste and aesthetics, I think are gonna be more important than ever before, especially with the rise of generated AI content um and knowing what's about to kind of take over our ecosystems in the next few years if it hasn't already to some extent, right? And so, how can we sort of shape like how can we learn from one another and like shape that and like through their program? I think would be really interesting to be a part of. Um, because what I also mentioned too is that like as marketers, we're taking our playbooks and we're kind of flipping them upside down to some extent. And so I think that this is a time to kind of rethink how we've done a lot of storytelling and marketing in the past, and which is why I'm like really curious to potentially sign up and be a part of it. Um, but yeah, so if you're listening and you are someone who, you know, checks some of these boxes, you should definitely look into it. I believe you have from now until the end of the year to apply. And they have a lot of cool speakers on their panel that you're gonna get access to, like Mark Andreessen, like Lenny. Um so I feel like they've got a really strong lineup of um, you know, fellows that you'll be hearing from during this program. Awesome. Yeah. Yeah. That's awesome.

SPEAKER_02: 1:32:45

I saw a post. Uh, I'm gonna summarize it. It was from Sanjay Blackdale G. I hope I'm saying that right. And the gist of it, I'm gonna scroll down a little bit, says for decades, the model was build a product, raise capital, use capital by an audience, paid ads. The world's top VC is now publicly admitting the capital loan isn't the moat it used to be. Tech isn't even the moat. The only true uh scarce risk uh resource left, pardon me, attention, that translates into trust. And then it goes on into really now it's that community is that that new moat, and uh this is teaching founders how to how to do that. So um I thought that was that was quite interesting. I think I saw that maybe Sunday or maybe it was yesterday. Uh I don't remember when I saw that, but I thought that was very interesting. So um Mallie wanted to get your your perspective on that. So um as we wrap up here, is there anything else you think um in the world of AI that's uh uh that's that's stuck out? That one from Jason Calicanus, that stuck out the most to me. And I was like, oh shit. Yep, that's Big Brother.

SPEAKER_00: 1:33:57

They're just gonna do what they stuck out, but it's not surprising, which is sad, which is sad because it's like this is where we are with these big corporate companies, not just in tech, but just across all these industries, is just such a a massive loss of trust because they're willing to screw over consumers in order to make their shareholders happy. And we're just seeing that, I think, across the board with so many insurance companies, I mean food companies, right? I'm not talking about just tech. So it's just like not surprising, but it's like big farmer, right? But it's just like bummer, you know. Um, cool, cool.

SPEAKER_02: 1:34:30

All right. Uh next week we'll be back with episode 27. Uh, we do have a guest on there. Uh, he's a badass. Uh, he was a very early investor in lovable. His name is Christian Peverelli. Uh Christian Peverelli. He's a builder. And he's got a sales tool that helps with uh well, CRM and sales. So it helps builders with customer acquisition. So I will have plenty of questions for that, dude, because uh I've been through that not in the AI space, but before when I sold ads, uh three time salesman of the year, thank you. Uh I had a certain CRM that I used, and um I I hate it. I couldn't stand the freaking thing.

SPEAKER_00: 1:35:09

So well, I'm I'm excited to chat with him too because I know that Lovable Today just announced that they've crossed 200 million ARR. Oh, and so as someone who's one year old, right? They're one year old, they just crossed 200 million ARR, which is it like absolutely unbelievable. And so I'd love to kind of hear his thoughts um from like an investor POV around yeah, what it's what it's like to be investing in a high growth unicorn startup.

SPEAKER_02: 1:35:36

That'd be banana. Like, how do we even get word of a deal like that? Like, uh, you know, here's a Swedish entrepreneur raising capital. Like, how do you even right place, right time, right connections?

SPEAKER_00: 1:35:47

I think I think it has to do with a lot of that.

SPEAKER_02: 1:35:49

Yeah, okay, cool, cool. That does it for episode 26. We'll be back next week. Peace. Bye everybody.

SPEAKER_00: 1:35:54

Peace.