# Meet the Hidden Agent Builder Turning Simple Prompts into Profit

Channel: Greg Isenberg
Video: https://www.youtube.com/watch?v=t3q5hCv_Kyo
Duration: 41 min
Language: English
Words: 8062
Transcript page: https://viewrankai.com/tools/youtube-transcript/t3q5hCv_Kyo

---

[0:00] You've seen those complicated NAND workflows that are all over the internet, but then you go into NAND and you find out that it's super super difficult to create these workflows. There needs to be an easier way. And I found it. It's called string.com. It's in alpha right now. Not many people are talking about it. And it's one of the easiest ways to deploy AI agents using yes, prompts. You can literally just prompt it and it's going to create those complicated workflows for you. I brought on the founder. This is the tutorial where he shares four workflows. The fourth one to me is the most interesting. Uh but it we go from

[0:41] easiest to most complex and I can't wait to see what you end up building. [Music] This is actually an emergency pod. We got Todd on the pod because he just launched string.com. Uh, which is kind of my dream product. Todd, if people stick through to the end of this episode, what are they going to learn with string.com? They're going to learn how to build AI agents that, you know, make their jobs more efficient or give them superpowers. Then that's the goal. Okay. Well, let's uh let's share your screen and let's get going. Um, all right. So, this is string.com uh which

[1:25] is an AI agent that builds AI agents. Um, and I I have kind of four examples. I spent about two hours this morning going through these. So, I tried to find use cases that were actually like real and would be relevant to a, you know, solo entrepreneur or a small business owner trying to like get stuff done. Um, and we'll just kind of go increasingly

[1:44] in complexity if that works for you. Cool. Let's do it. Um, so the first one I had was just, you know, build an agent that monitors hacker news for any mention of the term MCP. As you know, we we've built a lot of MCP servers. Uh, when new articles are discovered, you know, notify me in Slack. And I'll just go ahead and kick this off, but my assumption is um based on doing this this morning that it's possible there are so many articles mentioning MCP because it's like the the hot topic of the day that um the first version of this may be overwhelmed with volume. So,

[2:16] it'll be interesting to see how uh string handles it. So, the first thing I'll say, don't ask me this again, but um it always gives me a plan before I go ahead. This really prevents the tool from like burning your AI credits on like going in the wrong direction. Um in this case, it's a relatively simple prompt and so it should have success, but uh we noticed previously that sometimes, you know, the the AI would go build something wrong and then you'd be out of credits. You'd be like, "This thing doesn't work." Um, so essentially there's two tabs of string. There's the progress, uh, which basically tells you

[2:51] like what the agent is doing on your behalf. Um, you know, it's saying like, I'm going to go ahead and build this agent. I need to figure out how to get the articles from HackerNews. I need to figure out how to integrate with Slack. Um, for those of the audience who know Pipere, that's the company we run. We've got tens of thousands of tools available to agents. So some of these things uh will be pre-built and some of them will be dynamically written with code and that's part of this increasing

[3:18] increasing complexity I'm going to show. Um so here we are. We've got call of a simple agent. It the trigger is basically looking for articles that match the term MCP and it's going to send the message to Slack. It's asking me to test um and this is basically it could fail. It could say like it's

[3:35] somehow just not getting the articles. Let's let's see if it does. So, what's cool about this is you've sort of abstracted away um you know the the NADN flow charts that you know we're so familiar with nowadays and it's really now just prompting using plain English to actually create that. Yeah. I I actually went back to a slide that I presented in January uh today when I was just thinking about um you know why we went down this direction and it was really like if you can get out of this beautiful mind of like these really complex node charts that like everyone gets lost in and you simplify the

[4:13] interface to natural language which for us is English. I think you could make the product 10 times easier to use. And similarly, if you remove the dependency on this registry of triggers and actions, which we have one of the biggest in the world, but you sort of included codegen, then I thought you could solve 10 times more use cases. And that was basically the the drive to create um to create string. So Todd, I'm literally sweating just because this is exciting me so much. I'm literally going

[4:42] to have to take off my sweatshirt. That's a good sign. You know, that's a good sign. I gota be by the end of this episode. I'm I don't know if I'll have any clothes on. Um, strip automation is a new game. I haven't played it before, but um, so this is part of the magic of the product. So, what you're seeing here is it's asking me to configure my Slack account. So, I went ahead and said, you know, it's my pipe email and then it should ask, you know, what channel I

[5:08] want to send it to, um, as a next step. But again, this is kind of like built on, you know, five years of app integrations um uh and approved clients. So, I'll just say let's do Todd test. And so now those are really the only like config components. Um I've obviously connected my Slack account before, but if not, I would have went through like an OOTH flow. Um but my hope would be that, you know, once this goes end to end, we can just open Slack and we'll we'll see a message. Um, and once it tells me that we're good, I will go ahead and open it. That was

[5:44] successfully sent a message. All right, let's go see. Um, all right. So, this is sending, you know, per new article. So, I think it was relatively intelligent. So, instead of trying to send like 500 articles, it just sent me the most recent one as a test. And of course, if I deploy this agent, it will go ahead and send every

[6:03] new article mentioning MCP to me. Sam Alman, the co-founder of OpenAI, just said that it is the era of the idea guy and he is not wrong. I think that right now is an incredible time to be building a startup. And if you listen to this podcast, chances are you think so, too. Now, I think that you can look at trends uh to basically figure out uh what are the startup ideas you should be building. So, that's exactly why I built ideabser.com. Every single day you're going to get a free startup idea in your inbox and it's all backed by high

[6:40] quality data trends. How we do it? People always ask. We use AI agents to go and search what are people looking for and what are they screaming for in terms of products that you should be building and then we hand it on a you know silver platter for you to go check out. Um we do have a few paid plans that you know take it to the next level. uh give you more ideas, give you more AI agents and more almost like a chat GBT for ideas with it. But you can start for free ideabrows.com and if you're listening to this, I highly recommend it. I'm going to go back. It should say,

[7:15] "Yep, yeah, the agent is ready to run and we can deploy it." So, this is kind of like a basic automation, but I think what's cool about it is like I have someone on my team who actually looks for news and sends it to our product managers, product designers to just keep them up to date. Like that's a person that's

[7:34] actually going and doing that. Yeah. This just replaced that part of that person's workflow. Yeah. And I I love starting with like the most basic, which is like trigger action pair in our language. But when I built this this morning, I immediately said, "Can you summarize the article? Can you do sentiment analysis? Tell me if it's positive or negative. Can you write a clever comment that I could respond to the post that shows I know a lot about

[7:59] MCP and have a perspective on it?" Right? And you can imagine this just getting more and more complex, but again, instead of that like beautiful mind of like node craziness, um, you know, you're just iterating through through English. Absolutely. All right. All right. You want to go to uh one another example? Let's do it. I'm not promising that I'm going to take off my t-shirt, but we will see. Okay. So, I tried to find one that was a little like relevant to maybe um people who are content creators. So, I said, you know, build me an agent that monitors an RSS feed. In this case, I'm just going to

[8:32] use the Pipeam blog for new blog posts and then use AI to write like a viral LinkedIn post and put that content in a Google doc just so I can go grab it and and post it quickly. Now, you probably do this like I do. I I did give it a little guidance on the LinkedIn post. You know, I just said, "Hey, there's three viral components. Keep it short and punchy. You know, it should be one sentence paragraphs." Of course, if I was doing this, you know, for real, I would probably spend a decent amount of time on like the LinkedIn prompt just because having done this myself over a

[9:03] long period of time, um you know, the better your prompt, like the better the output and I would hone that over time. Um but again in this example I'm trying to now include like a few different things like do some some AI uh integrate with another app in this case um you know Google Docs and go from there. So I'm going to this is just one of the little nuances but I've turned on auto testing so as it builds it'll just automatically test so I don't have to constantly click test test test. But if you were doing an agent that might like delete rows from a Google doc, you know,

[9:38] you kind of might want to have human control over that. Okay. And what's by the way, we never talked about pricing. Like how how does this work? Yeah. So we we're really following the model of the vibe coding platforms, which is you basically have a subscription to the product and it comes with a pool of AI tokens. Um I believe the token count is 20 million on our like basic plan. And you can see I have 58 million tokens in my account right now. So um we can monitor through this process like how

[10:08] many I I use doing all these use cases. U but of course like our goal is to you know reduce the cost of token creation and token usage and just you know as these models get better reduce those costs. And I see a button for open in pipe dream. What does that mean? Yeah. So, um it's funny. I was uh I was in a meeting kind of demoing a very early version of this to Gummo from Verscell and he said, you know, my sort of view of these vibe coding platforms is they're best when they're built on top of like dev tools that have been built

[10:40] over like four or five years. And the reason is because you can have this escape hatch. So, we have this button where you can like open this into our old product um and essentially go back into that wizzywig editor and um you know make changes and do the drag and drop and all that experience. Um I actually slacked my team when I was doing this this morning um and I said I

[11:04] only had to do that once this morning. So to me it's like it's a little bit of an antiattern, right? like we want users to never have to open um in pipe and and eventually that button will go away. Um and I think that will be you know when we're no longer an alpha product but we're you know a successful product that

[11:24] button will be gone. It's interesting by the way I don't know if you caught that while we were talking but um it looked like oh it's happening again. Looks like there's there's a there's a some sort of error in the trigger. Um, I'm not a 100% sure what is happening, but it's obviously not receiving a trigger event. It should recover. Um, oh, here it is. It's explaining. So, it says it's it's not receiving data for the 25th event. So, it's basically like looking for blog posts that happened. I guess that was 3 days ago. Clearly, we need to produce more blog content as a company because

[11:56] there was nothing recent enough. So, I assume what it's going to say is I changed the date. Okay. I've set the published after date to May 1st. So now it's like it's going further back to looking for like the most recent blog post, which I think is actually pretty intelligent, right? It started with a narrow date range, then it probably expanded from 1 day to 3 days, and then it's like, all right, there's nothing there. I'm going to go all the way back to like two months ago and grab your most recent blog post. It' be interesting to see the the most recent blog post was uh June 11th. So it needed

[12:27] to go back, you know, a little more than than two weeks uh to find that. So I again I consider when whenever you're vibe coding and something fails but recovers that's more what we're focused on today really than than anything else is like trying to figure out how to get people to value and does it fail often and like how should people think about dealing with the failures. Yeah. I mean I think it really depends on the complexity of the use case but my sort of general belief is if you're just testing these tools you should probably get you know 50 to 75% if it's something like a prompt on

[13:03] the homepage should be higher than that but I mean if you're just doing like simple things want to see if I can get to value like I would think over 50%. If you're doing things complicated whether this is like you know replet bolt lovable or you know a product like string you know I think you're in that 25 to 50% success rate today and that's why these products are you know in our our product at least is alpha um you know like that's the percentage we're monitoring internally is like how do we get the majority uh of people to value um so it's interesting so it says it you

[13:35] know created a used open AI in this step to create a LinkedIn post. Um, one thing I'll mention is we we have this kind of like what we call batteries included for AI. So, we used to include all these AI tools, but you constantly had to like input your API keys or figure out should I use this image creation thing or that image creation thing. So, now when anything touches AI, we default to the tools we think are best for your use case and we just include our own API keys and then we just bill you credits

[14:06] when you run these agents in the future. It's just another one of those examples of can we get more people to value. Um so let's see if we created a Google doc. Okay. See um make sure this is a let's see um how alter builds integrations faster without having a single back. I mean again I'm not going to go through all the detail on this but like looks like a you know kind of like a legit LinkedIn post to me. probably would need to edit it, but um a summary of this

[14:36] this blog article. Yeah, it's insane. Yeah. So, and and if you wanted to and when you say edit, like you mean a human being edited editing it? Well, I'm just saying when I do the final link, I don't know what you do on LinkedIn, but like when I do the final LinkedIn post, I kind of take what the AI has given me and then I, you know, I polish it. That's like the last five, 10 minutes. I want it to be my voice. Um, and of course I'm trying to train it to get better and better, but I always think that last polish is like is

[15:04] the best human use of time. So what you could could have done here, for example, is add another step that pings a Slack channel that says the Google doc has been created and is ready for review, right? Yep. Absolutely. Yeah. And if you wanted to do that, you would just type a message. Like how would you how would you do that? Yeah. Yeah, I mean it's just like it's almost like an employee like your AI, you know, everyone's

[15:28] hiring these AI automation engineers. It's like the new hot like job wreck. Um, you know, we thought about calling this product just like your superhuman AI automation engineer, you know, because it's like working for you at all times and and available for for work. Um, I'm just not sure that that's a widely understood job yet, but maybe it will be eventually. So, what's it saying here? Uh, so it says the user wants to add a step to process when the process is completed. So it needs to configure a Slack notification step. So here it's just saying send Slack notification. Um, and then I assume it's got to write

[16:04] that. And then of course it's got same exact flow we just went through before. It's got to figure out where you want that to go. We've tried things where we just like automatically select Slack channels. Like so we have to be smart about it. Like there's certain apps where you can, you know, make an inference and figure things out, but it's like you probably don't want this, you know, little test that you're working on to exactly like click the general channel. Like you probably don't want to like blast your company um you know, with all your little tests. Um so anyways, that's that's kind of what it's

[16:34] doing. It's just figuring out like where to send it. Um and you know, I assume uh we'll see what it says. See, uh it's running the test. All right, successfully test. See here. I don't see Oh, yeah, it is. Yeah, new LinkedIn post generated. Nice. It thinks it's it's high quality enough to get a rocket ship emoji. Um, it didn't link to the document, which of course I could ask it to do. Um, but, you know, it's interesting. Here it goes. It gives you the document ID, but didn't give you the URL. So, and that's a case where I think it was kind of a dumb decision. Um,

[17:08] could have been better. We'll tell string to do better next time. Love it. Um, you want to move on to another example? Let's do Let's do it. Okay. So, I'm going to give an example. So, I was really inspired by um your boring marketer team that had like a little chat about string in, uh, on Twitter, which I saw, which is super cool. Um, I liked it because it was a real world use case. They basically said, um, you know, give me a daily summary of my, uh, Google Analytics. So, I'm going to fire this up and then I'll talk through it. Um, and so I would say

[17:40] this is an example where based on my test this morning, so I might need to jump to one that works. I'm thinking like 50% success rate. Um, and so already I got some ideas on how to fix this. But about three weeks ago, we sent an email uh it was like to a million people and it included a URL to a Verscell app and we basically built a demo of like it was chat it end up we ended up putting on chat.piptream.com pipe.com, but at the time it was a just a Verscel URL and somehow Verscell URLs like get flagged as like really bad in Gmail and these other email services. So

[18:17] like for the last three weeks I've been logging in every day and like into Google Postmaster tools and figure trying to figure out like are we out of the penalty box? Are we out of the penalty box? We out of the penalty box. So, um, I basically just said every day summarize, I should have said the last seven days of Google Postmaster statistics and send them to Slack so that I can and break it out by day and include a summary of the week so I can figure out, you know, basically are we essentially out of the penalty box or not. Uh, it turned out about, you know,

[18:44] a little over maybe four or five days ago, we kind of like got out of the penalty box. Um but the reason why this is particularly interesting is what you can see on the screen right now is you know we've integrated Google's APIs but we have nothing in the registry related to like fetch your Google postmaster tools. So this is all like dynamic code gen. This is the second half of what I talked about which is one communicate with natural language instead of like a wizzywig editor and two you know get out of the registry and write all these things dynamically. And so um basically string is writing custom

[19:21] code to hit the API to figure out you know stats and then we'll ultimately format that message for Slack and then send that message to me. Maybe we've overused Slack in this demo but we do live in Slack. Um, and when I did this this morning, you know, one time it it sort of veered off and got a little bit lost and it was really struggling to get

[19:40] the statistics from Postmaster Tools. And then the second time I did it, it had one error, it fixed it, and it worked. So, be interesting to see um, you know, how how successful the test is. I mean, if if this is able to work, like that's a huge that's a huge win, right? Like that's saving you a ton of time. Yeah. I think for most uh what I would say is like most people just wouldn't do it. Yeah. You know, like when I've talked to the CEOs of other um no code platforms, they always say the number one problem they have is their product is too complicated. And these are the

[20:18] people who run the least complicated noode tools in the world. So, I think that just the fact of the matter is if you have to go through some multi-step config that requires you to understand an API, you've lost, you know, at least half and probably more like 80% of the potential audience uh for these types of products. See, I already see by the way this is this is already I already know it's going to error because it says it's received data for seven zero days in the past seven. So, be interesting to see if it recovers um for that. Yeah, I think like as you're going through a lot of

[20:50] these examples, I'm just thinking about my own business like what are what are automations I can do to help save time and also just like make more money. Yeah. And I the way the way we like view this, we have like this worldview of these automation, you know, agents. The first is tasks. Like if you just want to send an email, send a message to Slack, add a row to a Google sheet, I think ultimately that's going to be built into the AI chat apps. You know, I think you're going to go to Claude and you're just going to be like, you know, like send a message in Slack and it's just

[21:23] going to do it. Um, so I I think that like you don't need a standalone tool for that. The second category is automations, which I think we're we're starting in automations in this uh conversation. We'll kind of move into true agents, but automations are things we all know. they're like zaps or pipe dream workflows or you know whatever the tool of choice is for for a company and I think those can absolutely move to a textbased interface. I think we've sort of proven that. And then the third category is what we internally call like real AI agents. These are things that have like autonomy and access to tools

[21:58] um and sort of don't need this stepby-step guidance as much as there might be a step that says like you know go do these things and it just figures out is there steps is it one step is it one app three apps um and we're spending most of our time you know on the product side either trying to get people to value just like the chat interface or building the the foundation for these more autonomous uh capabilities. I mean, that's the dream. The autonomous agents is the dream that you set a goal for one of these invisible employees and they have some constraints, but they're basically working at the problem and

[22:35] they're trying to get closer to that goal and they're they're working for you autonomously. And we really try like to not overpromise those things because I feel like half these tools are like it's solved and I'm like it's not solved, right? It's like it's there's a lot of momentum and green shoots and uh potential but for the most part and you know we have another half of our business by the way where we sell infrastructure to other AI agent companies and we see you know there's a lot of promise but it's early you know it feels like the first out of the first inning you know not even the first

[23:07] inning so all right so um anyways this one just said summarize postmaster stat um it actually had one test failure um and then it actually got to success and I said, "Hey, the key metrics are all blank." Um, and then it went ahead and and fixed it. Um, and then, you know, it sort of explained to me like the visual representation, you know, basically said, you know, here's your results for the last week. This was 8:30 this morning. All the metrics are good except for this one, which I know, which was the one where we got hammered on our reputation. Um, did a little bit of

[23:41] analysis and then said, you know, focus on improving sender reputation. And it was kind of like a, you know, no Um, I got to work on that. So, uh, but that was the, uh, that was the example that worked this morning. You know, this gets me thinking that like productizing agents and automations and then kind of wrapping wrapping that up and selling that to other companies is a great business to start. You know, I'm talking about like a cash flowing business. Like if you're, you know, I'm sure someone would pay $29 a month for that service set up. So, have you seen people basically start using string to create

[24:18] some workflows and agents and then selling that to you know other companies? Yeah, I mean we've we've talked a lot about I think it's a little like at least on our platform it's a little early but I I think if you look at what's happened with NAND and all these you know they call it the like the automation bros on like Twitter and and YouTube you have this demographic of people that you know maybe were in you know drop shipping and then we're in crypto and then are like they're in the business of making money on the internet and they've sort of discovered this like AI agent uh category as a potential

[24:50] market to like you know build products and sell them. Um, at least from my conversations with customers, I think most of those is more about today, at least generating views of their content and actually less about selling like quote unquote the templates. Um, but I think that there's a huge opportunity once these templates like are actually easy enough to implement for the end user that like I think there could be really big businesses built in these. It reminds me a little bit of like the kind of like the website template world where um you know the challenge is always can you actually get to value if you

[25:24] purchase it, right? And so these things need to be easy enough to implement for the end user and then I think absolutely you could um you could build a business on that. Before we get to the to the next one, do you have any you have one or two ideas for people if they were trying to build a big business on top of string? What what they what could they build? Well, I don't I don't think we've built the tooling yet for people to build a business. And I was going to get a little like little tech nerdy, but like there's really no reason why every

[25:55] agent on string couldn't also be an MCP server. And we we have a view that you know there will be MCP server businesses built where people are like here's an MCP server to I don't know like pick random examples process refunds on this SAS app or you know write blog posts in this way or you know do this other job function that people have. Um and so you know we've talked a lot about once we can prove that we can create like scalable agents at you know that are solving real business problems uh for companies there's no reason not to enable people to just like turn that

[26:30] into an MCP server that you then sell and market. Obviously we want to build tooling there so like it's like a oneclick and you you know submit it to marketplaces. Um but I I think there needs to be more demand frankly on the MCB server side. like it sounds good but like you know we need more people consuming them but that feels to us like that's the biggest business opportunity for like an enduser of our product today. Cool. Yeah. Um All right. The um and I guess that's that's skipping over the obvious use case which is like we do have a lot of like contractors who just

[27:03] like sell automations and then build them on our product but I feel like that's not really like us enabling them as much as just like they use the tool. Um, all right. So, I'm going to go with like a little bit of a more complex one. Um, and in this case, like I was like I figured the first question you were going to ask me is like what are the use cases people should solve? Like this is the problem with horizontal products, right? You can do anything. I was like great. Like what do I do next? Um, so this idea was basically build an agent that runs every day. It sends me an idea

[27:36] for automation that I can do like in my business. Um, and I tried to create something that would actually create like an end consumable end product. Like it would send me an email that looked good that like was, you know, in HTML that like, you know, like I don't know, had an end product I could look at. And then I try to get even a little bit more complicated, which is I just wanted to create a link at the end of that email where I could just open it directly in string and have it build the agent for me. So I was trying to almost create

[28:03] like a complete end to end like send me an idea, I click a button and then it builds it. Um, and this is going to go much deeper into a few things. One is there's going to be a lot of steps that need to be accomplished. Like string can either jam a bunch of this code into a single step or have multiple steps. So, we'll see how that works. It's going to do codegen. Uh, it's going to integrate Gmail. Um, and I'm going to set up for auto testing so that uh it just kind of

[28:32] goes through that process on its own. Uh, and then I'll pause there if you have any questions on it. It sounds like for the vast majority of use cases, auto testing should be on by default. Is that right? I think so. I mean, again, it's the only risk of auto testing is if the agent is doing something you don't want, but like if it's sending me an email, I'm not too worried about it. Like if it's sending me a Discord message, not too worried about it. if it's, you know, interacting with my product database in Snowflake, I might feel a little, you know, less confident in allowing it just

[29:06] to, you know, go rogue. And I will say like, you know, I'm sure this is true for every vibe coding platform. We've seen some examples where it's done dumb stuff, right? It's not common, but it does happen. Um, so that that's really the guidance around, you know, auto testing. Same is true with human beings, though. That that is also true. Yeah. Um, so it looks like what we have so far we'll see

[29:31] is it's basically generating the idea. It's creating like the prompt that's then going to share that I'm going to open up in string. It's creating the email content. Uh, let me just see if we're okay. Make sure we're not then it's going to convert it to markdown. I assume what it's going to do next is it's going to um oh it says convert down convert markdown to HTML which is good uh because it's going to send an email and now it's creating a template and hopefully uh integrate Gmail and send it to me. It's like relatively quick. I'm glad you think that. I mean, every time

[30:03] I see a spinner, I have like a little panic attack cuz you you just like you want these things to be instant. But, you know, when it's writing code and it's like progressing through um you know, like ideation and iteration, it's like it it takes time, you know, it takes time. Yeah. Um All right, there it goes. So, it's it's got a send email message uh action from Gmail. So, that's good. I do think by the way like you know any one of these AI products that's like built on top of the APIs of like the foundational model companies like every one of these things going to be better like the way I think

[30:41] about it is this is the worst string we'll ever be you know it's like the way I think about it like if it's good enough to get to value that's awesome but it's like it's the worst it'll ever be in um in our use of the product. Well, yeah, that's why I wanted you on the pod also is just because I'm a big proponent of play with the tools now, get good at the tools so that as the foundational models get better and these tools get better, it won't be like your first time trying these tools, uh, you know, you'll have some of the reps, you'll understand the guard rails,

[31:12] you'll understand the constraints, you'll understand some of the best practices. So, I just think it's a good idea to get your hands dirty. Yeah, I think that's true. Um, and like if if you're good at using AI tools today, like you're the unicorn, right, in your business, in your company, you're hirable. I mean, it's just like every single CEO I talk to is like, "How do I get my company to be more AI native? And how do I get people to use?" So, it's like I mean, it's just like such a way

[31:44] to differentiate yourself. Totally. Oh, good. I was worried there for a second. So, it looks like we've we've progressed. I don't know. That did seem to be a little bit of a longer uh duration than I would have expected normally, but um it's going to go through and test all these steps. What's nice about auto test, I don't know if you've done this on the like Manis or any of the the Vibe coding things, but you know, a lot of the people that I think are best at using these tools, like they've got like four or five tabs open and they're working on like multiple projects at the same time. And

[32:16] so one of the things about you know moving to these things to be auto you know an auto test or anything else that like it just it'll just get to the end you know you can be working on something else you know it's like it's like delegating a task to an employee so you might have four or five running at one time and then you don't have to like worry as much about you know it's the waiting game. If you do run four or five at the same time does it go any slower or how does that work? They're single threaded on our end. So like um you know

[32:45] if you have three or four in your account it's not uh it has no impact on the speed of your right experience. You're not you're not throttled or anything like that. No. Um interesting. All right. So we got a a test fail on converting markdown to HTML which is not totally surprising. So it'll be interesting to see if um if string

[33:05] recovers. It's an emotional roller coaster just vibe coding and vibe marketing, you know, because with the fail the tests and the failures and then it works. Yeah. Well, you know, one of the we we talk a lot about these AI tools, at least internally, is like there's this like magical experience when you get something to value. You know what I mean? It's like my best example is the first time I used Bolt and I clicked the

[33:32] button and then it showed me a website. you know, that was like the hit. That was like the the dopamine hit, you know, like it worked. And what's kind of interesting is I've found that when you hit errors and recover in some ways, sometimes you get a little bit more of that like I'm making progress. Like it's

[33:48] I'm getting that hit again and again. Um, you know, so it's um I don't know. It's been my own kind of vibe coding experience. Okay, so it says test failed then it tested successfully. Yeah, we could kind of like try to go in and and uh see what happened. I'm wondering if um yeah, it looked like it was just like a, you know, bad code on doing that transformation. It looks like here it did. Does look like it got it. So, this is just kind of like a little preview of the text. It looks like it got to success. Um let's see if where we are

[34:22] now. Sly sent chat. Okay. Um let's see if it if it gets all the way through. You know our hope of course is that the user never has to look you know at like you know JSON or output or open and pipe dream. That's the goal right um you know today the more sophisticated you are and the more technical are obviously you'll be more powerful at these tools than than non-technical people right. Um okay so here it is. It's it looks like our prompt is is going to be something

[34:57] about seamlessly onboarding new clients. Um let's see with slow manual onboard process loss revenue. Okay, it's like pitching. All right, we got one more error. Let's see what we got. If we don't hit this one, we um Oh, interesting. It's going right to Gmail. So, let me go ahead and just see what we

[35:16] got out the other side. I just used a demo Gmail account. So, my gut is given we had one error, I guess I'm skeptical to it's going to work on the first try, but we'll see. We'll see. Mhm. You're right, though. When it does, you know, when you do fall and recover, it does give you that dopamine hit. Like right now, I have

[35:39] that uneasy feeling in my stomach. The slot machine, but you you haven't gotten the results yet. Totally. I'm like getting anxious. I'm like, "Oh my god, is this going to work?" Yeah, I'm uh I'm skeptical, but let's see. Um let's see if this All right, we got something. Let's see. Okay. Um it's not actually that bad. The the thing the error I was getting earlier this morning was it was like not converting the markdown to HTML. So it'd be like this kind of like garbage. All right. So today's automation idea intelligent client onboarder orchestrator. Um, so it's basically saying, you know, many businesses have slow onboarding

[36:17] processes. Imagine end automation that transforms clients. All right, not not too bad. Um, it's going to do this automation. It's giving some ideas. Probably don't need to read through all this. Um, here's the generated prompt seamless and then I'll I'll just click here. Let's see if it works. It's gonna seamlessly onboard new clients with automated workflows. Now, I mean, if I click this, it's not going to solve my problem. It's a little too generic, but I'm pretty happy. Like it went end to end, you know, it got something like I wouldn't be embarrassed if you know this email went to like a customer or an internal team, you know? I mean, like

[36:54] for I don't know how long that was, 10 minutes, you know, not too bad. Yeah. And I think like this is step one, right? Like you can make it better from here. Yeah. So I think that's a part of it, right? Yeah. I love that it's selling competitors in the in the email about Pipe Dream.

[37:14] Bad string. We got to work on that. It's just it's uh it's doing the best it can. Yeah. Well, again, these things are based on historical data. So, it's like it's interesting when you launch a new product like string, I always wonder myself like how long before, you know, open AAI is going to recommend this in chat GBT to customers, right? It's like it could take could take months, you

[37:37] know? I it depends on how good they are. And I don't know if you're seeing that in your business, but like I checked last week and now it's like a third of the traffic to our website is coming from AI, you know, chat bots, you know, rather than Google. So it's um it matters, right? These new products are, you know, important to get in the system. Todd, thanks for coming on. If people want to get started, well, first of all, I should make make mention I'm not involved in string at all. You know, I just invited Todd on to to show this product. I think this is an interesting

[38:08] product. I'm starting to use it. I'm trying to get good at it. And that's why people I think like listening to the show is I bring on interesting founders building interesting tools to help people save time, make more money and propel their startup ideas, you know, in into reality. Um, so I wanted to just say that. But Todd, if people want to use string, what is the best way to get started? like you know it is a bit daunting to look at a blank screen what do you want to automate so do you have any you know I guess advice for people yeah well the first thing I would say is

[38:44] thanks for having me on the podcast I'm a big fan of yours and and consume a lot of your content and I appreciate kind of what you're doing for entrepreneurs and small business owners and so thanks for being thanks for having me here um I I would recommend starting with like everyone has in my opinion one to two hours of just operation ational BS they have to do every day in their job, right? It's like the things that people generally like doing the least and it's like the kind of grind of being a knowledge worker. And I always say like just start there. Start with something

[39:18] you spend, you know, 15 minutes a day doing or an hour a week doing or a few hours a week doing. Don't try to create an AI salesperson that like closes deals for you as your first agentic product, right? It's just like automate something that you have to do that you don't like doing that would save you time or money and then you know kind of iterate and get more complex from there because otherwise I found that people they dive right in the deep end something doesn't work and they go this stuff doesn't work and I don't think it's a string issue that's just a vibe coding vibe

[39:52] automation whatever you want to call it issue so it's like start smart small solve a real problem get to value and then you know you'll understand at that what edges you can push in terms of complexity and and value. All right, man. Thanks for coming on. Appreciate you. Awesome. Saturday podcast only from now on. Yeah, I told Todd so we're recording this on a Saturday. I was like, this is an emergency podcast. I was texting with him last night. I was like, I need to understand string, so

[40:21] you got to come on. And uh thank you. Thank you for spending your Saturday. And people uh like, comment, subscribe uh if you want more of this in your feed.

---

About this transcript
Read from YouTube's own caption track and laid out by ViewRank AI (https://viewrankai.com).
ViewRank AI finds the videos already beating a creator's own average on Instagram, TikTok and YouTube Shorts, transcribes them from the audio itself in more than 60 languages, and turns what worked into new ideas and scripts.
Free transcript tools, no account needed: https://viewrankai.com/tools
How to read any video this way: https://viewrankai.com/llms.txt