We were thinking long and hard how not to over-query XRPL full history nodes, and came up a solution. Data sharing!
Each instance of analyzer can be registered to central cluster. If user requests account info "xABC..." from your instance, normally if you don't have full history of that account
on-hand, XRPL Full history node will be queried. On the other hand if you are part of Analyzer cluster, other instance could provide you with data dump of account "xABC...".
That data dump is already analyzed Ledger data connected to "xABC..." account and it can be imported directly to your instance way faster.
Analyer consists of lots of small PHP packages, think Symphony or Laravel, they are frameworks which are built with smaller packages.
XRPLWin will provide specific packages which can be used independently, such as XRPLInterface, Liquidity parser, XRPLUtilities etc...