REST API structure at a medium sized project?

Hey everyone I was wondering what does your basic REST API controllers look like?

I miss the rails way of

• dry.rb based form validation object
• using AR directly in the controller for saving
• rendering response

Instead of the current elixir way:

• create make a changeset in the schema
• make an update function in the context for that particular controller method
• connect the schema to the new function in the context
• use the context to validate, update and save

How do you all do it? It feels like a lot of extra steps and a lot more functions that rely on each other.

Also with this pattern, I’m wondering how this would work at scale. I’m already feeling the chaos between polymorphism, 3 features, and 5 different background jobs working around polymorphism.

I noticed that theres the (drops)[Elixir Drops 💦 — drops v0.2.1] library which feels like dry.rb which is nice, but something feels off on how to manage context, schemas, etc at scale.

Would love to hear some ideas on how y’all manage this. I’m missing rails a little bit in this aspect, not going to lie.

I know I should ‘stop thinking OOP’ and think that it’s all functions but it feels so abstract. I need a way to organize the functions and the default way feels quite overwhelming. I’ve read the DDD book but this paradigm coding and organization feels like I’m spinning a web of complexity (even though this feels right).

Anyways I know this is a bit of a rant but I’m getting concerned of the future of expanding this project that is slowly growing more and more into something that feels unmanageable.

Any advice?
How do you all manage complexity and things like these?

Still feel so new even though I’ve been writing elixir for a few years now…

Thanks,