Web Development

I think you're using the word "framework" quite broadly here. Cloudflare and databases aren't "frameworks". The word is perhaps ill-defined, but generally it means something like "an opinionated library ecosystem in a programming language". React, Vue, and the like are Javascript "framework" libraries for making what are called Single Page Applications. They are hardly the only way to make websites, though, and some (including yours truly) believe them to much more trouble than they are worth for most purposes.

Laravel, Django, Ruby on Rails, etc. are all backend/server-side "framework" libraries for creating Multi-Page Applications, or in other words, the model for which the web was originally designed (that is, you get HTML and follow links embedded in that HTML to get to new pages).

At any rate, a "file system for a website" is quite vague. It really depends on what your goals are. You have to define what is you're trying to do and who it's for. It's not particularly difficult to spin up an sftp server to allow remote access to files if all you care about is remote copying, but I'm going to assume you mean a file server served over HTTP(S).

There's about a million tools out there for serving up static files, getting a directory listing, etc., and it really depends on your goals. It's fairly simple to configure nginx to serve up a directory with a very rudimentary index if you don't need anything fancy (nginx can do fancy too, ofc, just takes more work than what comes out of the box). I personally like Caddy over nginx these days, though. Caddy in particular requires very little configuration, honestly.

But yeah, without more information about what is is you're trying to accomplish, it's difficult to give specific recommendations.

I'm not quite sure what you're looking for, but the term "non-binary data" is sending me haha

Anyways. I've used Laravel for about a year now, and it's ..fine. I don't like the push towards AI slop or their constant attempts to push their own cloud services or whatever onto you either, but it hasn't been much of a hassle to just ignore that part.

What I like about Laravel is just how opinionated the framework and the documentation is (like how it wants you to build Factories and set up proper test data when making new database Entities, and how the Inertia routes encourage a clear schema). It's also been pretty pleasant to get the basics right out of the box, like proper auth and login functions. And I like the kind of stuff that you can do with blade templates! I'm using them to serve dynamically generated SVG files and it works really well.

I haven't really needed any sort of crazy file management things though, so my opinions might be completely irrelevant to your use case. In fact, my Laravel project is really not set up for any sort of large data, since that's just not a thing I need. I'm running it in an AWS EC2 instance (the smallest one, as that's included in the AWS Free tier) with just an sqlite database right in there. If I suddenly wanted to manage files in my application, I'd probably be looking at doing that with AWS S3 buckets. Which would work I guess, but I personally really prefer interfacing with S3 buckets via AWS Lambda, and those work much better with Node, Go, Ruby or Python. I'd probably be leaning towards Node, since that's just what I'm most familiar with. Doesn't mean that it's the choice I'd recommend, because your project might have completely different technical requirements! Web dev is a confusing clusterfuck of edgecases.

If Laravel looks like it works for your requirements right now I'd say go for it. No one knows what the future looks like - it could be completely enshittified in a year, or it could be totally fine because the maintainers have just been riding the AI hype for marketing clout while still making a decent framework. If you stick to decent coding practices, switching to a completely different framework really won't be the end of the world.

Alternatively, since you're looking to hire someone for your project, maybe just focus on finding the most capable engineer you can and let them choose whatever tech stack they seem fitting? They'd probably have a much better understanding of your requirements and could make choices that work well together contextually.