Stop Overthinking Your Tech Stack

Key Takeaways

How do you decide on the text stack you should use for each project? Here's the answer from my friend Sha Talibby, a builder and AI engineer, previously data scientist who shipped tons of products and is working super efficiently. And the answer might in fact be simpler than you think. So one thing I'll do which I mentioned earlier is I'll like chat with chatb if it's a big project try to get some clarity and then I'll take that to uh Claude to you know start thinking through how to start implementing this like what to implement first and things like that but as far as like tech stack I try to keep my tech stack very simple and just follow the principle of building with what I know. So I just use Python for everything. Um, I will typically use OpenAI's API just because that's what I what I use uh most often. And really the the place I'm at right now is very much like the zero to one part cuz really the projects that I build other than being like tutorials for the educational stuff is I'm building like little AI SAS products. And so for that really the goal isn't to make like the most efficient and scalable and beautiful piece of software. The goal is to just build something quickly and try to validate it as quickly as possible. So that's why I just try to keep everything super simple, you know, modular, composable, things like that, but also, you know, just kind of optimize for speed just so I can get it out into the real world and start collecting feedback on it. As Sha pointed out, when you want to ship fast, the text stack isn't really a priority. You need to focus on things you are most comfortable with, which usually includes more simple text tax and as important having a modular and composible approach that you can reuse from project to project or further expand on if the project is working. You don't want to just do a proof of concept that isn't scalable.