Steps to Add Multi-Provider Functionality

[M123-dev]

Hi everyone,

First, thank you for this great project! I've been exploring the codebase and am really impressed with how much functionality is already in place. However, I’ve been looking into extending the repository to support multiple providers and wanted to get some discussion going on how to approach this.

From what I’ve seen in the code, it looks like the current implementation, especially around the PlaidItem model/class, is tightly integrated—Plaid is currently the only provider supported. This makes it a bit challenging to envision a straightforward path for adding other providers without significant refactoring.

First, the elephant in the room: it is interesting to know what is the vision of the team is:

• Lots of different providers in the repository (locked down)
• Extensible add-on system (similar to Home Assistant?, any third party repository can be added if wanted)
• An API where anyone can write a little cron script to add things (like Firefly iii)
• Something completely different?

Besides that:
Provider Abstraction. Is there a plan for abstracting providers, or would this need to be developed from scratch?

[zachgoll]

@M123-dev great question. Currently, to prioritize speed of development we've intentionally gone down a highly-coupled implementation to Plaid. This will be the case for quite some time so that we can focus heavily on achieving a high data accuracy rate with a single provider before moving to a multi-provider flow.

That said, the long term vision is what you describe—a multi-bank-provider abstraction where our users can choose their preference (or potentially, even add providers as "plugins").

Since we can't support multi-provider abstractions quite yet on main, the best path to implementing this would be an independent fork for the moment.