pinpoint-apm/pinpoint ? reverse-engineered prompt

Build me a self hosted application performance monitoring system for large scale distributed apps, like Pinpoint. I want a web dashboard that shows service topology, live traffic, request and response trends, slow traces, call stacks, error analysis, and machine stats like CPU, memory, garbage collection, and thread activity. It should work with Java applications and collect data through an agent with as little code change as possible, and if you can support other common runtimes too that would be great. Please make it feel like a real monitoring product, with a clean admin UI, searchable traces, and a way to drill into one request and see where time was spent across services. If anything in the docs is unclear, look up the current Pinpoint setup details online and keep the build close to the standard project structure so it can be run locally and tested easily.