Incomplete builds, unreliable systems, undocumented code, and applications that outgrew whoever built them — diagnosed honestly and fixed, or replaced, without pretending the original approach was fine.
Before committing to a rebuild or a fix, I go through the codebase, the infrastructure, and how it actually fails, then give you a plain answer: what's salvageable, what isn't, and what each path costs. That's a Workflow Assessment applied to an existing system rather than a manual process — see pricing on the home page.
Stop the bleeding first — the bugs or outages actually costing you money or customers right now.
Write down how the thing actually works, so it's never a single point of failure again.
Fix what's fixable. Replace what isn't. Either way, you end up owning something you understand.