Ask HN: What data synchronization tech should a new web project opt for today?

I'm starting a web based project (actually my first) so I have no history or technical preferences yet I'd have to consider. The project can be imagined as some sort of collaborative scene graph editor, so I have to synchronize a tree-like data model among browsers and the server. Regarding choices for specific libraries I'll be going to use, the data store and synchronization aspect seems to have the biggest impact on the general application architecture. What modern, well documented, flexible technology/libraries would a well-informed web developer have to choose from today? I have no need for backwards compatibility, but I'd prefer as little dependencies as possible and if that's an option today I'd also like to avoid transpilation. One important aspect to me (although I'm not planning to implement it from beginning) is offline use.