One thing I’ve learned over the years is that good architecture often starts with understanding the system you already have.
Before changing something, I usually invest some time looking at how requests actually flow, where the dependencies are, what the failure points might be, and perhaps most importantly… why things ended up the way they are.
Sometimes there’s a very good reason. Sometimes the reason is simply that the system evolved over time. Either way, understanding the context usually leads to better decisions than jumping straight to a solution.
A small lesson I learned a long time ago, but one I keep coming back to.