I'm new to the world of virtualization. Please help direct/forward my post if this is not the best/right forum.

My project will use x86 with KVM/libvirt in an embedded platform. Initially, I'd use an evaluation platform with Intel XeonD processor. I could use virsh to prototype/experiment with various virtualization features. However, I'd think that whatever steps that I go through with virsh would eventually turn into invocation of some libvirt API in production level code.

How do people typically translate a virsh command into a libvirt API (1-1 mapping)? Is there another tutorial or web page describing this? Or is my approach flawed?

There is extensive documentation on language bindings here: https://libvirt.org/bindings.html