Build AI Agents in Minutes With Lindy (easy)

If you had to fire 80% of a marketing team today, which 20% would you keep and why?

That is such an interesting question. I think I would keep the managers. I think basically what AI agents mean is like we all got a promotion, like we're all managers now, like, and who we are managing are not necessarily humans, they're AI agents. So I would keep the managers and I would keep the builders. I would keep people who in particular are able to manage these AI agents. It's going to perhaps self-serving. But, you know, we are seeing more and more of that internally ourselves, like our marketing team, the average person in the marketing team spends probably at least 30% of their time just building and managing agents. And that's up from basically 0%, like 2 years ago. And I think this is only going to go up over time.

Yeah, that's great. So we're here today, guys, to talk about the uncomfortable truth about AI marketing. And Flo's got a lot to share. He's the CEO of Lindy, which is a tool that we, a product that we pay for that we love. And I've used it, I think, in a very basic way just because I'm kind of jumping all over the place. You know, I'll jump to Replit, I'll jump to Cursor and all that. But I think it's a very important staple in our kind of AI tool stack. And so Flow's got a lot of cool stuff to share. So mutual friend of ours, Greg Eisenberg, did a pod or more so like a YouTube video. And we're going to go through a lot of cool stuff that Flow's working on today. And I'll probably ask some follow-up questions, but Really, this is gonna be for marketers and also business people that just want to grow faster, acquire more users, grow faster, and gain more efficiencies. And I'll probably jump around and I'll have some follow-up questions maybe afterwards too. But Flo, do you have some stuff that you're ready to pop up and then I can just, you know, I can ask questions and poke at you?

I have stuff. As we go through it, Eric, in particular, if I share my screen, can we like cut and like edit afterwards? Oh, for sure.

For sure. And I think it's also important for you to maybe describe what Lindy does too, just so everyone knows and just so they can start to understand the possibilities.

Yeah. Okay. Awesome. Yeah. So for quick introduction, Lindy is the world's easiest and most powerful way to build AI agents. We often describe ourselves as the Macintosh of AI agents. It's like every technological revolution goes through these phases where it's like phase 1, it doesn't work. That was like AI agents in 2023. Phase 2, it works, but only if you're a nerd, like only if you can code. And so that was AI agents in 2024. Phase 4, phase 3, sorry about that. Phase 3, it works and it works for everyone. And that's like the Macintosh for PCs, right? So that's what we're driving towards. We want to make it super easy. You shouldn't have to be an engineer. We want to make it super easy for you to be able to create AI agents.

Just so everyone knows this, so by the end of this video, you will learn how to build simple AI agents that actually drive your business forward, not just a bunch of, you know, pie-in-the-sky stuff.

That's right. Yes. We use, obviously we use Mindy all the time. We have thousands and thousands of customers who use Mindy a lot, like millions and millions of tasks a day. So yeah, no, I mean, this is for real. Let me share my screen and show you some examples of how we use Mindy for marketing. So I'll just start with, Comment magnets. So we have this LinkedIn post here that we posted five days ago, actually. And at this point, like we're posting one of those every week, and every time they go viral. Like this one, look at that. It's like three thousand comments. It's just like a lot of comments. And so in this one, what we're saying is like, hey, we built an agent that tells you exactly how to sell to anyone. And like comment agent, and we'll DM you the agent. And then like just people commenting agent, agent, agent, agent. And that is a LinkedIn post went viral. We obviously can't DM everyone like that manually, like 3,000 comments. It's basically a full-time job to do that.

Mm-hmm.

So what we did, and this one is really interesting because famously LinkedIn does not let you, LinkedIn does not want you to build any automation on LinkedIn. And I spoke with someone very senior at LinkedIn about 2 years ago actually. And I asked him, it's like, hey, can you guys build APIs? Like we need more APIs. Uh, uh, because we, we really want to build more automations on LinkedIn. And he was like, no, like, we don't want you to do that. And I was like, hey, we're going to do this the easy way or the hard way. Basically, it's like, you know, it's like, you know, computer use agents are coming. And at the time, it was a hot take. It's like, computer use agents are not coming. It's like, it's totally coming. And if you don't let us do it via an API, we're going to do it via computer use. And sure enough, this is where we are today in 2025, um, where we have this agent that Every day wakes up, looks at the latest comments, and then goes on LinkedIn, looks at the profiles, and it checks at first if it can send them a message. And if it can't send it right now, it's like, oh, you can't send it because this person has locked in their DMs unless you pay us. And so instead, you can see the bot here, and you can see on the right what it does click by click by click. You can see the agent go back in the comments. and type a comment and mention the person and then tell them, hi, I would love to send you the template. Feel free to DM and I'll get it to you. Because if they DM us, then we can DM them back. And under the hood, it's super easy the way it works. Whoa, there you go. It's really super easy the way it works. And we just made a release like a month ago where you no longer even have to create these workflows yourself. You can just like prompt it in natural language and it's going to create it for you. Um, but it's basically, um, where is it?

That's a big call out, Flow, just real quick as you, as you pull it up too. So like earlier before we started recording, I'm just like, yeah, you know, I've been using, you know, you know, different workflow tools. Like, you know, obviously the one everyone knows would be Zapier like 15 years ago, right? And then like Flow is kind of like, you know, it's, I think this is the evolution of workflows and things like that. But, um, I think what you mentioned was important because you just launched recently natural language where you can just like say, Hey, you can just use SuperWhisper or WhisperFlow or something and just dictate out what you're trying to build and it will build these workflows out for you. And then maybe you can just edit these as you see fit, right?

Yeah, 100%.

Yeah.

And I think like WhisperFlow is actually a great way to do it. We also have a dictation feature, but you would be shocked. It's way better to like ramble and talk for like 5 minutes to this AI agent than it is to like type a couple of words in there.

Which is fascinating. Go ahead.

Yeah, I mean, like the way it works, like super simple, it's just like there's an AI agent here that like, you know, you start the computer, then we inject in the context window the messaging templates. And then it's like, okay, first you check if we're connected to the person or if we can DM them. And if we can't DM them, then you go to the comments and you send them a DM to tell them to connect with you.

Cool. And so that post alone, I think I saw 1,000 comments. So I have to assume maybe 500 of it Marvin's responding to, but 500 is still a lot. It's really eerie.

Oh, 100%. I mean, this is a lot of comments here. And I mean, that post alone here, I mean, you know, we would need to do the math, but like probably earned us like multiple hundreds of thousands of dollars in like 5 days. Wow.

So what you guys are doing right now, your flow, no pun intended, your flow here is that you're just sharing cool agents that you guys have built and that everyone wants to build these things right now. They're very simple. And then you have this comment agent that uses Computer Use to DM them back. What are you using to, you know, do you guys, are you creating these landing pages from scratch? Because that's the problem with these, right? So like, I'll do a post and like, you know, it'll go viral, get a bunch of comments on it and things like that. But then it's like, man, you have to create a landing page and then they have to opt into it and things like that. So how are you getting over that? Like, that's another annoying marketing problem.

Yeah, we actually have another Lindy that creates It's not exactly landing pages in this case, it's more like SEO blogs, but basically anytime we have a new use case that we see start taking off in our data, what we do is that we kick off this agent and then we create like 10 or 20 pieces of content in like 5 minutes and we send them. And so again, this one's like a really simple workflow. It's like, okay, you receive a message from a user with a request to write an SEO optimized blog post. Okay. Then you use Perplexity to analyze, uh, the, the, the website that you want to write this for and the topic and the keywords that you want to optimize for. Then you write the blog post and then you send it back. It's, it's that simple. Um, and, and, and it's at the same time way more powerful than just asking ChatGPT to write this blog post because it's going to be anchored in actual research.

Got it. And, and, and so Flo, this, for this one, so it's creating a blog post, but there's no opt-in required for this whole comment thing. So you guys aren't trying to collect the lead?

Um, well, I mean, once it's created, I'm not sure I understand the question. Once it's created, we just sell it on Webflow.

So, okay, I guess just tell me if I'm not understanding this correctly. So you guys are making this, okay, here's, here's the agent post on LinkedIn, dude, I created this new agent over here, comment if you want access to it, right? Typically, once someone comments, they might get a DM saying, okay, go to this landing page over here if you want access to the agent, and then you have to opt in for it, and then they're on your list.

Oh, I see. I see. I see. I see. No, all we do is like, we just create these templates. Like we have actually created the agents that we talked about in the post. And then we just turn this on right here and then we copy the link and we send them this link. And that's just like a template link to an agent and they have to sign up to access it.

Oh, so they have, okay.

Yeah.

Got it. Got it. It's like, here's a setup and then it's like, yeah, I mean, if you're going to proceed, you probably need to sign up for Lindy. So that's how, that's okay. That plugs in together. Um, that makes a lot of sense. And like, I just want to emphasize everyone here that If you're doing this, the comments piece to Flo's point is a full-time job, but making these pages is super annoying. And sometimes people say, oh, I'll do it in a couple days, I'll do it in weeks. Like, now you get it done in minutes, right?

Yeah, yeah, yeah. We are also creating a Lindy that's very good at getting meta. Like, we are currently creating a Lindy that's going to create Lindys for us, and that's going to make these posts for us on all sorts of social media, including Instagram and TikTok and YouTube Shorts. Like, it is also going to generate a video with like using AI of someone talking about these agents. It's like, I just created an agent that does X, Y, and Z. And the whole, the entire thing, like the creation of the agents, the creation of the video, the posting on social media, the like comment magnet handling and all of that, like 100% of that is going to be handled by AI. So that's why I led the way I did, which is like the job of the future is just to manage these like— Hundreds and hundreds of agents. Yeah. Dude, this is good.

I'll let you get a sip of your coffee, but like, I definitely want to see more, but I think it's, it's, do you have any interesting numbers you can share around what you're seeing with the growth of agents overall on Lindy or whatever interesting things that will make people go like, wow. Um, any numbers you're happy to share around your business?

Oh, I mean, you know, we're growing super fast. Like we've, we've like much more than 10x over the last 12 months. Uh, I'll just leave it at that. So. But it's also like, look, it's, it's, uh, I think we've got the wind in our back. Like, you know, we were in hindsight early in AI agents. Like we started working on the agent, uh, pre-ChatGPT, like mid-2022. Um, and, um, you know, it's, it's just, it's just the technology is now ready and it's like, once you see it, it's like, oh my God, like it's just, it's just crazy. So people, people are very excited about this. Cool.

All right. Let's see some more. And then I want to ask you more questions around how your feelings on remote work in general, 30% of time doing AI agents, AI fluency and all that. There's a lot for us to talk about, but let's see more agents. This stuff gets me excited.

Yeah. Oh, same here. Okay. So like right here, this is another example. This is an agent. We also have a YouTube channel and we also actually follow YouTube channels that talk about us. And so we have this trigger here that's like a new comment was posted in this channel. And when that is the case, when someone comments on the YouTube channel, it sends us a message on Slack. And so for example, here it's like, why is this template not working for me? And so we receive a notification in real time on Slack, and then we can actually respond on Slack. Like, we can interact with this agent fully on Slack. Like, I think that's the other beautiful thing about AI agents is like, they really meet you wherever you are.

Huh.

And so we can be on Slack and we can be like, hey, tell me more about the video that they commented on. And so here's the agent is going to invoke the skill that it has to transcribe the video. Or like, can you please research this thing that the user is talking about? And here the agent is going to go on Perplexity and research the thing, you know? And then we can just reply to the comment right here on Slack. So like that lets us be super responsive for these YouTube comments.

I have a question, Flo. So it's interesting because you're choosing Perplexity for the comments here and also for the blogs too. What have you seen in terms of difference? Because typically what people will pick for writing, they might pick like Claude for something, Or Claude. Why Perplexity?

Well, Perplexity is really just the research layer. It's just really a way for your agent to pull information from the internet, but then the agent is the one that's going to do the writing. And yes, under the hood, it is also powered by Claude. Like, you can define—

Oh, I see. I see.

That's right. So that's one big thing that we're really believers in here at Lindy is like model agnosticism. So for any step of any of your agents, you can decide which model you want it to be powered by. I like Claude 4 Sonnet best. Like, that's what I use for most things. Sometimes you might want to use Claude 4 Opus for like really intelligence-sensitive workflows. Yeah. Great. Yeah. We also do a lot of influencer marketing. I think like in 2025, if you're not doing that, that's a problem. So we have this other agent here. What it does is it helps us find influencers in our target market. So one second, let me try to open this. So here it's like, hey, I want 20 influencers on YouTube with 20,000 to 40,000 subscribers in the productivity and AI space. And then it's just like, it's using like Perplexity and a bunch of different tools. It's doing a lot of different searches. Then it's cycling through all the matches that it found. And it's doing extra enrichment about all of these matches. And then it logs all of that in a spreadsheet. It's like, okay, I have found these 7 people. You know, this is what I found about them. This is their focus. This is like how you contact them. This is like, they've already got some brand partnerships in some cases. And I've added them to the spreadsheet.

I love it. Yeah. So I think in this situation, I mean, it sounds like at the very top, it's like if people are getting confused at all right now or overwhelmed, I don't think you should be because you're just basically saying, yeah, you're using Super Whisper, Whisper Flow, you're having a— you're rambling for 5, 10 minutes or so into, you know, into the prompt, right? And then it creates this workflow for you and then you just continue to edit the workflow. And is that how your team is making these right now? Like each day you just think about all their problems and then they just start to dictate it out and they just start to create more and more workflows?

Pretty much. Let me show you something. So this is, this is an example of, of, of a workflow I created to like manage my personal CRM. And yeah, I mean, like the way I created it is like literally I just prompted it, help me manage my personal CRM.

Hmm.

And then it's like, all right, I'll take care of it. Uh, and it, you know, it asks me some questions like, hey, is this what you want? I'm like, yeah, that's what I want. And then, and the video is sped up. Like it doesn't actually take 10 seconds. It takes like 60 seconds. But then I had this simple workflow here to manage my CRM. It's like, okay, if you want to add contacts, that's what they do. And if you want to retrieve contacts, That's what I do.

Yeah.

So, um, yeah, it's, it is very easy now. One of my pet peeves, and I'm sure like your listeners, uh, dislike this as well, there is so much hucksterism right now going on in AI. It's just, there's so many grand promises of like, look, this is it. This is the AI employee. Like, look, you know, it's, it's early. These technologies are in beta. And I don't want to pretend that you're going to be able to automate your entire marketing function from just a prompt. Like, that is not the case. You are going to have to work at it and look, it, It does take a tinkerer, right? It does take someone who, you know, if you've already used like a workflow creator, like a no-code workflow orchestrator platform before, that's going to be much easier for you. But once you have learned to use this kind of platform, they get incredibly powerful and they can make like a real difference in your business. And we see that every day across our customer base.

Got it. Can I, if we play around from my side, maybe, so I'll share my screen real quick. And because we have the Lindy CEO here, Maybe you can give me thoughts on, so I've had this workflow for the last 4 months or so. Okay. I want to upgrade the workflow, right? And I'm trying to figure out how I can do that through natural language. And so maybe this is like a demo to the team to see how this might work with mine. Are you—

We can do that if you want. If you've tried like the natural language workflow creator, again, it takes like 1 or 2 minutes to create a workflow, which like is pretty short when you're working. It's quite long on the podcast.

Like we may need like some elevator music or something, but yeah, I'm happy to run the Yeah, I think this should be pretty quick too, and then I want to come back to yours, and then I want to go through the practical sides too, because if people are like, oh, I can't get my team to figure this out, but you guys clearly have, so there's got to be, I want to know the flow process for this. Okay, so let me share this with you real quick. Can you see my screen? This, I think it's this one. Here we go. Okay, you see this? Okay, so this is our daily standup accountability bot, right? And so the beauty of this is this actually coaches people. So it acts like a, it looks at the 3 things you worked on yesterday and then the number 1 thing you're working on today, just to make sure you're working on the highest leverage stuff, right? I launched this maybe 4 months ago and then instantly it made everyone prioritize better within a week, right? But I haven't had time to continue on this. And some people are like, oh my, most people are like, oh my God, I need this, right? But then some people are like, oh my God, you're policing your company, blah, blah. I'm just like, dude, we just want our team to perform better. That's all it is, right? Now, what I would say is, okay, I'm here now. I'm like, the next level would be I wanted to have memory and look at the last quarter or so, and I wanted to tie in with everyone's individual KPIs, the company goals and all that, and I want to help people continuously improve, and so this is a coach that continuously helps them improve for each quarter. And so literally what I told you, I'm like, I can probably just dictate that out. I'm in this task section over here, but how would I upgrade this agent?

So let me understand what you're wanting it to do then. Like what do you want it to remember in its memory?

So all it's doing right now is it's looking at the last 24 hours of chat, right? Ideally, it looks at the last 90 days or so, and it looks at each individual and it helps that individual get better according to their goals. So right now it only looks at the last 7 days of raw raw chat or last 24 hours. So it doesn't have memory. It doesn't align with their goals either. So it's just constantly giving surface-level feedback without really understanding the individual and their progress.

I see. I see. Can you go to your flow editor? I'm curious how this is set up. Okay. So this is like actually a very simple agent. You know, so memory is one of the least solved problems right now in AI agents. Like the easiest way to go about this is simply to edit the memory skill. So if you go to the agent step, um, and you edit a skill using the plus button at the bottom, uh, no, no, no, no, sorry. Uh, at the bottom right, like in the agent step configuration.

Oh, sorry, sorry, sorry.

Yeah, yeah, yeah.

This one? Right here.

Yeah. And, uh, I think it's called add memory or modify memory. Right here. It was called modify agent's memory. Boom. Beautiful. And then give it the instruction, like at the bottom of the agent step prompt here, like the purple text field.

This one?

No, no, no. On the right, like the agent step, like above. It's a purple text field. Oh, this guy. At the bottom, just say like, hey, after you're done collecting your information, or like, I'm not exactly sure how it works. Update your memory with like what people have been up to. Basically tell it what to remember. Right? So like, this is where you really tell it. And also you can update your prompt anywhere else in the agent step to tell it how to use what it's remembered before. Okay, cool.

So I'll test this in the future, but it's good to see this in real time. And I mean, maybe, so I'm going to stop my share, but I think this ties in with how Like, how is your team using, because when you say that 30% of your team every day they're making new agents or spawning new agents, like, how is that, like, how have you set up the AI fluency expectations? How should people learn how to use this?

Well, you know, in our case, we're cheating because we're Lindy, so like we just live and breathe the product every day. But I think like what just happened is a good example of what I meant, where it's like, look, it's not rocket science. It took you like 2 clicks and like a prompt to do it, but you do have to spend the time to learn the platform. It's like you kind of know that you have to go in the agents tab and add it to skill and let it memorize stuff. So we're not yet fully in the Macintosh of AI agents moment. You still do sort of have to know how to use these platforms. So I'll answer your question in a different way because how we have reached AI fluency is we just built Lindy, but our customers, the way they do it, is that most of the time you sort of need to crown someone like the czar of AI at your company. You really need it to be someone's basically full-time job.

David Sacks.

Yeah, exactly. And so it's scary for people because like, oh my God, it's a headcount and all of that stuff. But actually in my experience, it pays for itself in spades. Like we have a customer that did that. They have a czar of AI inside the company. And literally one of the things they unlocked for this Star of AI, through this Star of AI, is for the entire rest of the company to be able to make changes to the application, including the support team and including the product team without ever touching the code. So they've set up like Devin, like there's a lot of these like coding agents right now. And now like the product team can just go and be like, hey Devin, like look, this button here, the copy is not right. And it's like small things like that, but like in the end it saves you, you know, it's probably worth across the company like half or like one, one full engineer. So it already paid for itself right here for this AI doll. So crown an AI doll. Most of the time people already know in the back of their minds who should be the AI doll because every company's got this one person who can't stop talking about AI. And then give that AI doll the objective of like, hey, it's not just you setting up AI inside the company. You need to be a wind in the back of the rest of the company and you need to empower all of the company to adopt these AI tools. And I think today there is an opportunity to do it. Like, if you do it, I think you really have an opportunity to pull ahead from the competition. I think tomorrow is going to be table stakes to do that. And if you don't do that, you're going to be falling severely behind.

I love it. So are there any— so when you, for example, if we go to recruiting for a second, do you guys have any expectations with AI fluency? Like, is there any special question you have? So for example, for us, We just ask people just straight up, like, what are you paying for out of pocket when it comes to AI? If they answer $0, they're basically out of the running. So like, I'm sure you guys have something better than that. And what would it be?

You know, what's the saying? It's like the shoemaker's son always go barefoot. No, I don't know that we have like much better than that. Like, I think in positions where we're really like, the job is to create AI agents, then we just have an interview where we like sit down and create an AI agent and like, You can tell immediately whether someone's going to be able to do the job. Like the rest of the time, like we do interview for like curiosity and resourcefulness and intelligence generally. And in my experience, if you have these things, like you're going to have used Claude or ChatGPT, it comes out in interviews whether they have used them. If they haven't, then it's like not even worth speaking to. Like who, like what the fuck? Like if you're in 2025 and you're not really using these tools all day, like then you're You know, I think very behind.

Hey, I'm with you, man, but it's, I just, there's always anchors in a company I don't want to adopt. And you have the fanatics, you have the converts who will follow the fanatics, but then you have the anchors that will drag it all down.

So yeah, we may also be taking it for granted. I think we might be benefiting from a selection bias, like people who apply to Lindy are AI fanatics, like always.

Yeah.

Yeah.

Yeah. Okay. Um, now, okay. I want, I want to see if you have anything else to share on the, on the flow side in a second, but I'm just curious, like as Let's say I'm a lazy business owner and I want to hire Lindy consultants to help me set this all up. Do you guys have like a section for that?

Uh, we do have something on our website where you can find these consultants. There's also like a lot of those on like Upwork and like Fiverr and all of that stuff. So, uh, absolutely. We also offer those for like, uh, for higher, uh, ticket items for like, you know, larger companies, but we also have our own professional services arm internally that, that does that all day.

Yeah. Okay. That's good to know. So maybe I'll come back to that in a second, but are there any other workflows you want to share before we work towards wrapping?

Let me see. I think I talked about the influencer lead generation. I talked about the SEO one. Well, I don't know if I shared, like, again, this is part of what I mentioned, the fact that we are soon going to have the agents generate the LinkedIn posts themselves. And so one thing we're doing right now is like we have these other agents that generates LinkedIn content from the content of our meetings. So Lindy has the ability to join your meetings here. So she can record the meeting. Um, and after the meeting, we just like use the content of the meeting as like a LinkedIn ghostwriter and an X ghostwriter. And then it adds all of those drafts to a Google Doc and to Google Sheets. Oh, that's so good. Then we can just at the end of the week, just like check this list and like decide what to pick and what to post.

Well, you just made me feel really stupid because we have a Lindy agent that's called the Compelling Ideas Note Taker. So it sits in because. You and I, like, the best ideas or the best content is going to come from when you're in those meetings and when you're not thinking about creating content, right? And it rates it, but then I didn't think about taking it to the level where it just writes the post and then writes in my language and all that because I didn't want to think about setting up like all these new boxes and these next steps. Now I'm just going to dictate it out after this and I'd be stupid not to do it. So that's amazing.

Yeah, no, 100%. Well, not 100% you'd be stupid, but like, I think it's just like inertia, right? It's like people just go about their day-to-day lives, like they don't want to change their routine. But I do believe that at this point, like you should be spending at least an hour per day creating and tweaking AI agents. And that's going to more than pay for itself because like once you've created an agent, it's just on autopilot and keeps doing work for you in the background. So like you just do it and just compound. It does more and more for you every day.

Yeah. And why would you not want that? And I want to reinforce this just so everyone understands. Just so I'm clear, it's every single employee at Lindy spends at least 30% of their day building AI agents?

At least, for sure, 100%. Insane.

Insane.

Okay, cool. Which is also why you can find it in the bottom line of the company. Well, one, we are growing fast, but two, our growth is very efficient. And our investors and my advisors are always mentioning, wow, you're surprisingly small for your revenue size. And I'm like, yeah, it's because we use AI agents all day. Yeah.

Fascinating.

Okay.

So now the templates that you shared with me, like the question I have in my mind is like, dude, I want to do everything Flow's doing. Right. So, um, do you, cause I know you guys have some templates, some, some baseline templates, but are you guys going to open like a public, like templates thing where people can share everything that they've built and you can rate them over time?

Oh, absolutely. Yeah. So that, that's coming up. Uh, but right now, you know, we've got this really high quality curated list. Of templates because we actually used to have what you just said, like a way for people to publish their templates. And it wasn't the quality of the content that we aspire to put on the internet. And so now what we've done instead is like we have handcrafted our own templates for the top use cases and like we've made them super high quality. And so you can just go here and like on the homepage of Lindy after you sign up, you can see the marketing templates. And again, like here you can see, I think there's like 20 of them. Or something like that. If that.

Got it. So you guys are kind of controlling this right now, but I just imagine like, because some of the other ones that I've seen, when they shared, it's like the ratings are public. And so you can, it's kind of like a public thing where like, if you suck, you get flogged for it. Eventually, are you going to do that?

Probably, yes. Okay, cool.

When do you think that's going to come out?

I don't know, a couple of months. Yeah. Okay, cool.

Yeah, I can't wait to, because like the stuff you built, I'm just like, On one end, I'm just like, I just want to pay to access your templates. On the other end, I'm just like, well, no, I can just kind of dictate it out now. But like, that's how lazy I am. I just want to pay for your templates. So that's some customer development for you.

No, I think that makes sense because I think like if you have to dictate it, I think that solves the ease of creation problem, but you're still going to have to come up with the ideas yourself. And the reaction you just had is one that we get every day. Like people just watch how I use Lindy and how we use Lindy as a company. And they're like, oh my God, I hadn't even thought of that. Like I could do it if I wanted. I just need to see that for 5 minutes. I just hadn't thought of it. So that's the problem that is solved by these templates tools.

I feel like you guys are going to have a revenue arm where it's just like, dude, that sounds amazing. Can you just still just, can you still just do it for me? And you're going to have agents doing it for them probably. Right. So that, that's, that's what I'm just curious from a business standpoint too. You guys charge on usage right now because it's a credits-based system. Are you guys eventually going to add on outcomes-based pricing?

I don't think so. I think it would make everything so much more complicated. I think we are a horizontal platform. And so like the outcome is also very much dependent of like how you've set up your agents. And it's up to you. Like you can make your agents like much cheaper, much more expensive, much faster, much slower, like take more or less context. So we do want to leave that control in the hands of the users to like really control their own agent. Yeah. Great. Okay.

Final thing I want to ask you about right now. So, you know, I think, and I talk to my CTO about this all the time where it's like, yes, my company, at least it's remote, but we used to have an office, right? So what is your philosophy on remote right now? Like what's your current setup at the moment?

I have done a, like famously a 180 on remote. Like my previous startup before Lindy was called Teamflow and it was like a virtual office for remote teams. And so I was one of these guys like banging the remote drum so hard because At least on paper, it's a panacea, right? Like you can hire from anywhere, like you save money in office, you save time in commute and all of that stuff. I think like, look, we've run the experiment with COVID on like a large scale and it failed. Like, and I think like the category of virtual offices also failed. Like, you know, like we were like flat and like doing okay, but like it did not explode in the way that we expected. And so I've totally done a 180 and I just think like, There is no software substitute to physical colocation. And I think that if you are optimizing for creativity, you want people to be together in person. You want to hear the music together. If you're optimizing for efficiency, then sure, be remote. And by the way, inside the same company, different functions may optimize for these different things. Like support, for example, is not a creative function, especially once you're at scale. It's like, hey, I just wanted today it's like $2 per ticket and next year I want it to be $1.50 per ticket, right? Like that's just an efficiency play. And sure, like send that to the Philippines even. Like it's like more than remote, like, you know, BPOs and so forth.

Yeah.

I think if you're optimizing for creativity, and I think people underestimate the extent to which building software is a creative endeavor. If you are optimizing for creativity, you have to be together in person. It's like the Beatles. Like you're not trying to make the cheapest, it's like, oh guys, like can we do it for like a buck 20 per song. Like, imagine how sick that'd be. Like, no, you're trying to make like the best possible song you can do. And in order to do that, you've got to be together in person. There's just no substitute for it.

Yeah, I love that. And are you guys hybrid, like 3-2 or just 5 days?

Well, 6 days. So hybrid, you get to work from home 1 day a week.

9-9-6.

Basically, yeah.

I love it.

Okay.

So this is interesting because my CTO and I, there's a period, I mean, we work out of an office. I'm heading to an office right after this, but There's a period of time where like I had a family thing happen, you know, he traveled summertime and then there was 2 months basically, maybe 6 weeks or so, we weren't in office and we came back, we're like, dude, us in the office, like we completely destroy, annihilate the remote situation. And so our whole thing is like, we want to hire more out of LA. It's not to say we're going to eliminate remote completely, but, you know, we were doing hybrid way before the pandemic and I'm just like, dude, you just, you cannot beat the magic of, uh, in person if you're trying to do a creative endeavor. Here's the problem with most marketing agencies: they don't understand what is happening with the world of AI. With my agency Single Grain, we think about agentic workflows, we think about creating agents for the future. When we think about SEO, we think about creating different products such as internal linking products or content consolidation products. And on the other side of things, we think about building products such as Care. We just think about making marketing easier over If you want to stay at the cutting edge of marketing and at the same time leverage the power of AI, go to singlegrain.com and we'll see you on the other side.

100%. And I think people talk always about like, oh, but you know, it's so much cheaper, like San Francisco is so expensive. And I think it's penny wise and pound foolish. I think, again, if you're optimizing for a creative endeavor, like your job is not to do it on the cheap. Like it's kind of like making like a Hollywood movie. Like if you're making a blockbuster, it costs you $200 million and you make the best possible blockbuster that you You're not trying to do it for like $200 or whatever. And so if it were true that doing it on the cheap was the way to go, then you wouldn't see the vast majority of successful startups come from the Bay Area, which is by far the most expensive place to build startups. And I think it is one of these cases of, there is this French economist from the 19th century called Frédéric Bastiat who wrote this essay that's called The Seen and the Unseen. And I think the perpetual mistake that people make in business is to sacrifice the unseen on the altar of the seen, is to sacrifice that which cannot be measured easily on the altar of that which can be measured easily. It is very easy to measure like, hey, I'm paying this engineer $70,000 a year versus $200,000 a year. That is so easy to measure.

Yeah.

It is really hard to measure the chemistry and the quality of the work that comes out of the team, but that's what matters. That's real. Like that comes out, right, of your business and that's what drives the business.

100%.

So yeah, don't sacrifice on that. Just like pay what it takes to build the most fucking amazing products that you can.

I'm with you. Nothing great is built in comfort. All right. Last question for you, Flo. Is there anything else that you think I should have asked you that I might've missed?

We didn't talk about AGI.

Yeah, let's, let's, let's talk. I mean, what do you guys got going with AGI right now?

Honestly, I have, I have nothing original to say about it. It's coming. It's exciting. It's scary.

Yeah. Elon tweeted today that he thinks they can— Grok can actually hit AGI. So I saw that like an hour or two ago. So yeah, we'll see what happens there. But Flo, I mean, this has been great. I mean, we'll probably have to do this again. I'll probably have to match every time Greg does one of these just because I want to selfishly, I want to continue to see the business use cases and the marketing use cases, but what's the best way for people to find you guys online?

Yeah, well, like on Lindy.ai, I'm quite active on Twitter. We're active on LinkedIn.

All right, Flo, thanks so much for doing this.