Transcription
I'm going to show you absolutely everything I would do differently if I could start my project from scratch. Now, this is my project. It's seogrowth.ai. If you're on either Shopify or Woo Commerce, feel free to sign up. We are also doing WordPress very, very soon. It's not just going to be e-commerce soon. It's going to do absolutely everything, but it's basically an agentic system for growing websites.
But one of the problems I have is that it was actually vibe coded from the very beginning before I even knew what I was doing. So in this video, I'm going to give you everything I would do differently or everything I would do now in order to start my project and make it the best success possible. If you're interested in SEO, it's segrove.ai. It'll be one of the first links in the description of this video.
So the first thing I would do is I would choose a back-end platform. Now, this is possibly the most important decision you will make. Uh, to be honest with you, you need front end and back end. So, you do need to think about this a little bit. But previous to what I used to say, I probably would use a template or framework because frameworks are just better for security. So, you could use, for example, fast API if you want to use Python, PHP just as its own backend if you want to use PHP and then next.js although I don't think next.js is actually back end. I think it would be node but I just put Nex.js just so people kind of understood.
And then although you do have to it says don't worry about front end here. What I mean by that is don't worry too much about design and things like that at the very beginning. Work on getting your brain fixed before you worry about what it looks like. So don't start off by worrying about your front end. I personally would just create an entire system of scripts that do exactly what you want and then you can very easily just plug in a front end to that. Right? If you try and do both at the same time, if you try and code front end with backend, right? I'll tell you right now, your your back end will suffer. Okay?
So, the next thing that I suggest you do is choose your payment provider and set it up locally. The way that you do that with Stripe is actually extremely easy. Just ask Claude Code to set up Stripe for you. Make sure it has access to the CLI itself and then make sure it's actually running Stripe locally so that it can test itself. And then what you want to do is you just want to swap out the M variables for the dev variables for the prod variables or the main variables, the the actual variables for Stripe. It's actually very very easy. This might seem complicated, but it's extremely easy. Once it's got itself set up, you won't have to worry about testing because it can test itself, right?
Choose your database provider and set it up locally. This is extremely important. I used Superbase, but you could also just use something like SQL light with permanence, which would also help a lot. So now, what I suggest you do once you've done all this is you create your entire brain system. Focus on your brain. Make sure that it works really really well. Make sure it's got all the AI done, etc., etc. And just make sure it's really really working.
There's actually one step missing uh from here. So, let's just say F here. Set up docket. Once you've got your AI systems um created, etc., and your designs created, then I would suggest that you get a real dev or an AI dev to stitch everything together instead of just trying to do everything at once. It just won't happen as easily as you might think. So, E is actually set up Docker. Docker is so important for this process. You do need Docker. Just get Claude Code to set up Docker for you. Get it to set up your fast API, whatever you choose for front end. um it can more easily run for your payment provider and your database and everything can run on Docker. So definitely set everything up on Docker. That's super super important.
So yeah, like I said before, get a real dev or an AI dev to stitch it all together. Obviously, this is not ideal, but this is just what I would do personally if I was trying to make another production SAS from scratch. And then finally throughout the process keep the code simple commented and read through everything Claude does. And then one more thing which I talked about briefly in the last video was foundational vibe coding or stepbystep vibe coding. What this basically is is you start by doing database, then do I don't know Docker, then do uh payment provider. The entire time you're doing stuff, every time it codes something, get Claude to update the claw.md with what it's done and also what it's doing. Right.
This is a new kind of thing that I've been looking at recently where when you start a project instead of just jumping in and doing stuff, you think about things from the very beginning. Now, this is a fundamental kind of programming concept that has kind of been lost in the vibe coding source. But what people used to do is when they were like approaching a project like Grove, one of the first things they would do, and my friend Ed taught me this, he's he's a developer, um is they basically get all of the tables uh in a database and they define the tables. That way they know the entire scope of the project because they know how many tables are needed in the database to create the project. And this although this can change the database is obviously they can change it in the future. It will pretty much look as it was originally designed in the final project and this is just something that is not done anymore in Vive coding for example by people like me who don't really know what they're doing with code.
So one thing you might want to do is you might want to plan out your entire database. You might want to plan out your entire structure. Put all of that into a claw.md file. start to implement it bit by bit. Another thing you can do is have an MVP document. I've been finding this pretty useful. So probably like around this stage here after you've done all the basics, you just say, "Okay, now this is the MVP document. I want you to just work off this MVP document and implement as much as possible."
So this is kind of what I've been doing more recently. I think honestly with this methodology, I could create an entire SAS in under 24 hours. If people want to see that, I will actually stream that live and we'll see if it's actually possible. I will leave the video there, guys. Thank you so much for watching. If you're interested in learning a little bit more about this kind of stuff, check out the school community. If I just go on the classroom here and I go to, for example, beginner vibe coding concepts or vibe coding with Claude code. There's a lot of the stuff that I just talked about in this video in this classroom here and also cheap vibe coding and also building production ready apps creating a solid foundation for a project. This is basically everything I just talked about in a bit more detail. um and then also building aentic systems which is another thing that I've been looking at in more and more detail recently. I'll leave the video there guys. Thank you so much for watching. Check out the school community. It'll be one of the first links in the description as well. If you're watching all the way to the end of the video, as usual, you're an absolute legend. And I'll see you very, very soon with some more content. Peace off.