I Built a Fake SaaS Startup with AI in Under 2 Hours

Okay, so what you see here is something is an experiment I've been wanting to do for a while. So, let me explain a bit before I get more into it. So, basically uh if you have been hanging out online like the last couple of years, at least since AI kind of took off, I've seen more and more of these posts like I bootstrapped my SaaS to 10K MMR in a month, and now I know what to

do. I used AI. I did this in a week. I have 10,000 users. You can see this on X like MMR 50K. I did this in like week, 100 days, no ads. And they get a lot of engagement. They put up these nice graphs and stuff like that. So, of course, most of these are fake, right? So, the what I wanted to do is an experiment to see how effective is this, right? Because if you see

them everywhere, there has to be something, right? So, I want to run this experiment uh to see actually what the effect of running these kind of types of campaigns, if you want to call it that, is. So, I came up with a plan um So, this is what I want to do, right? So, the idea is to use as much AI as possible, maybe some AI agents, set up some automatic workflows or something like that, to

see how effective we can actually use this. So, my plan is to create like a slick landing page. Uh I have an idea. I'm going to talk maybe a bit more into it, but basically the the idea is going to be some kind of uh quant betting page startup you can join, right? And there's a waiting list like right? Waiting list. Uh this kind of creates this FOMO, right? And you kind of want to get in,

and you do stuff like that. And I want to create like a fake demo. So, this should be like a video. Uh I thought we can use cloud code like with the AI agent, Codex, and stuff like that. And for marketing, uh I I did think about the YouTube, but I might just drop it. So, maybe we focus on X. I have like an account there. So, we kind of focus maybe the demo on X. We

might do YouTube. Let's see, but uh basically, I think I'm going to focus on X. Today, I thought we can just time it. Time. Take the time. How long does it take me to set up this? And I want to see the effect of this. So, we're probably not going to see the effect in this video, but I'm going to come back with a second video on this and see how it went. So, today, I'm just

going to time this, see how long it takes to set everything up. So, yeah. I think we're just going to get into it. I'm going to do a quick explainer, and then I'm going to just get started, I guess. Okay, so, this is what I'm going to use. Uh I'm going to use this X account here. This is an experiment account I used earlier this year. So, I'm just going to maybe switch it up a bit,

founder of I'm going to put it something in, change maybe something here. And for videos, I think I want to use hyperframes because then we can kind of use the HTML style we have here. We're going to create like a branding package just using like um ChatGPT image, I think. Uh for the web page, uh I'm just going to host it on Vercel since I have an account there. And that means that we're probably going to

do this in like Next.js or something like that. We're going to just use Cursor, I think, just for the VS Code style to look at the code. Uh one instance of the GPT 5.5 medium Codex, Claude Code, uh Opus for 7, I think. And I think that's it. I don't think we need any more than that, to be honest, just to get going. So, this will probably voice over, and uh but I'm going to record it.

So, the first thing I'm going to do is just build out um just going to build out the web page, the landing page. We're going to generate a video uh with hyperframes and stuff like that. And I'm just going to do a voiceover of everything I do. So, some parts will be sped up and stuff like that. But, uh I'm going to time it. So, when I start this now, uh we'll see how long I'm going

to spend on this. So yeah. Think we just got to get into it. And I come by it now and then and just talk about uh how we end up with this. So, yeah. Let's go. So, I just started here on Claude code with this long put uh input prompt I did with uh voiceover. And you can see I spent 4 minutes on that. But, I guess that's a good start. Uh I also did use like

a reference style here. But, uh yeah. And you can see I went ahead. I bought the domain maxquant.com. It was pretty cheap. And of course, we needed like a logo. I used the image GPT image for that image too. I think it's pretty good to be honest. And I just uh yeah, tried to do some parallel stuff here and also create some asset maybe for our later tweets. You will see why. Uh yeah, I just used

the photo I had on like my existing Twitch account here and tried to create some I don't know if I want to use them. So, this was the first version of the site I got from Claude code. Uh didn't really like this. But, uh of course, with Agent super fast to rework this. I got my logo. I wanted to kind of use that. So, I think I just put that in the public folder. And also, we

got this uh polymarket background. Maybe I was considering using it. But, this was the second version. And you can see now we kind of have everything here. We can start our computing trade. And yeah, we have something to work on now. And you can see I wanted to actually link this up with uh hyperliquid. So, it kind of looks like our trade has like an outbound link. And this was maybe the third version. Looks better, but

I wasn't super happy. So, after 21 minutes, was it? I think something. Uh I created this. So, I didn't record it. I forgot to put record, but I wanted to create like this I've seen this a lot of these viral dashboards. So, I hooked it up to the to the Polymarket web socket, and I can just stream a lot of data, and it looks just looks sick on video. So, I really wanted to do a video

on this that I'm going to use in the tweet. You can see here. This looks pretty cool, right? And this really shines through on the video, and you can scroll down, and you can see everything. So, basically, I just copied this from like a tweet I saw. So, I prepared myself to do a video on this. And you can see here, if I put in like space talks, we get into this mode that's kind of calculating,

and this was the final version. That looks pretty cool right? So, this just get to 100, you get your trade, you can move on to to Hyperliquid. That was the whole idea. So, I I'm going to play a bit of the video here, so you can listen to that. X USDC on Hyperliquid. We are already logged in. We can make a trade right away. We can also follow up here and ask what kind of trade should

we do. You can see some statistics here. Just such a so much you can do here. So, this is the setup we have now, and this has been super popular among our users. If this sounds interesting to you, >> [music] >> you will find the waitlist on our homepage, because we don't [music] have enough places to serve everyone now. But if you're fast, join the waitlist. There should be room for you coming up soon. So, yeah.

This is the first [music] iteration of Moxquant. Hope you enjoyed it so far. Okay, so, yeah. That was pretty good, right? So, I just collected some more assets. Yeah. Pretty straightforward. Uh this was the second part of the page I created because we wanted a landing page. I embedded a video. I could set up everything like you can join the waitlist. Uh yeah, I think that was basically all I wanted. So, this was the front page,

the landing page with the waitlist. I just removed a lot of stuff, tried to refine it, and basically, that was it. So, I just pushed it to GitHub. Uh yeah. You can see I added some Yeah, I can't remember I did maybe title or something. Uh prepared my Twitter account that I was going to do kind of a marketing on the fake SaaS marketing. Set up on Vercel. I think I deployed successfully. Yeah, that was a

good. And from there, I just set up the database for the waiting list. I just pushed that. Uh yeah, you can see I prompt for the Neon SQL uh database for the waiting list. That's the only thing I needed. I And I created a tweet. You can see uh yeah, you can go to this You can see finally my startup has come through. Uh I just embedded a 2-minute video that's good for engagement. And I can

see I'm thinking, and I just posted it. And the site should be up now on like maxquant.com, I think. I had some DNS issues, but I think I fixed it pretty fast. Yeah, you can see here I fixed it. So, the site is live. I had some more assets. Uh actually, I wanted to try HyperFrames here to actually create a a second promo video. Uh I think I'm going to play some of it, but uh yeah,

updated my Twitter profile from the old account, and that was basically all I did, to be honest. So, and you can see analytics, we are up and running. And I think that took me I guess I'm going to show it soon, but if I remember correctly, it was around like 2 hours or something. So, yeah. That was everything quite fast. This is pretty crazy, right? So, it's been like almost 2 hours, right? And if you go

back to I guess you saw I did the X post. Think we are at like five view five likes or something. Five likes, 107 views. But I will try to blur this, of course. But you can see we already have one on the waiting list after 1 hour and 50 minutes. I did some pauses in between, but basically effective work almost 2 hours. And we already have a waiting list. So yeah. That was basically my whole

point. But the idea is that I'm going to come back to this video, this homepage. I like you saw me do in the Yeah, I guess the voice over. We did the homepage, we set up the signing list, we did everything, we bought the domain, we created I have a second I have a second Let me turn on the sound here. So, I did ask Hyperframes to create a promo video. It wasn't that good, but I

guess I can play it here. This is a Springline train from West Chillville. We are now arriving at Triplet Peaks. Please keep all bags, tails, and belongings [music] free from the doors. >> [music] >> Yeah, that video needs some work, but basically the idea is that we could automate this from now, right? Because we can set up this pipeline that kind of post on X here autonomously. We can set up like a Reddit pipeline that do

this post. So, I did schedule a post. I guess I can show you that. If you go to draft, scheduled, you can see I did this post here. We are already at and we added a MRR media like we talked about. Uh yeah, I added this image and of course I added a screenshot of the dashboard and just uh I would never see this coming. I tagged some companies. I I don't know what to say. It's

just very strange that it this works so good to be honest. So, now I guess I could just schedule a lot of um I could schedule a lot of posts, right? I think I'm going to come back at the end of the weekend and show you the second part what we ended up with. And yeah, basically my conclusion on this project. Because I think it's pretty scary to be honest. So, please don't go swarm at that

page if you see it. I guess you can go check out the page, but uh don't intervene so much on like X and stuff because I want to do it like organically just to see what happens. So, like I said, this was the experiment. Run like a fake SS company to farm waitlists to create FOMO and stuff like that around an interesting product. So, I've just seen this anyway and I kind of wanted to see the

effect myself of it because yeah, I think it's going to be a big problem. It already kind of is, I guess. So, we didn't even try that hard and after like yeah, 20 minutes we have someone on the waiting list. Uh but I had a head start though. I had like an old X account. But, uh, many people have that. So, yeah. That is the experiment so far. So, I'm going to come back on, let's say

Sunday. Uh, with the follow-up video. So, look out for that. And, I hope you enjoyed this so far. And, let's do the conclusion. And, I'm of course going to take that down after the experiment is over. And, I'm going to delete the wait list, of course. So yeah. Uh, hope this was interesting so far. This just shows how much you can do in so little time now with these agents. And, I didn't really feel I pushed

it that hard. Uh, but I was happy how the page turned out. And, it looks pretty cool. And, the video was super easy to make. And, yeah. So, yeah. Hope this gave you some inspiration. Not, maybe not so much inspiration, but hope this gave you an idea what is going on out there, right? Because, don't believe anything you see. A lot of these pages that report these crazy results probably have nothing. Most likely. That is my

idea. So, yeah. That was it for today. Hopefully, see you again on Sunday. And, bye-bye.