Using Neo4J to replace traditional DB in OO virtual environment

I've built "Hex", a distributed OO environment that feels like Smalltalk. The name is a homage to the late Terry Pratchett. I want to use Neo4J as a backend for Hex. I've been developing in this space since stumbling across Smalltalk-82 in 1982. Hex is a pure OO environment built in Python and built upon a virtual machine stack that allows it to operate in at least a few "interesting" languages (Python, Javascript, Lisp, maybe others) while finessing the "dot-net catastrophe" that results from attempting to force every language to use the same bytecode set. Hex assumes a cloud-based environment that is always running, has universal object identifiers, a write-once paradigm for changing things, and forgets objects instead of deleting them. For those of a certain age, think "SecondLife for objects".

I'm here because I find the documentation and startup materials for Neo4J impenetrable. While I'm a newbie to Neo4J, I'm reasonably familiar with Linux, virtual machines, networking, and so on.

I know that, like many newbies, I'm missing something blindingly obvious. I'm hoping that this community will help me achieve that first "Aha" moment. With any luck, I'll be able to do the same for others once I'm a little less disoriented myself.