Every seasoned engineer knows the tension between what the codebase needs and what the product demands. As systems evolve, legacy decisions accumulate, and the codebase becomes more complex, and more difficult to maintain. That's when the urge to refactor, or start from scratch comes in. This roundtable invites experienced engineers to discuss around the real-world complexity of refactoring at scale: when to do it, how to argue for it, and what to do when it’s simply too expensive. When does “ugly but stable” code become a real liability? What are the hidden costs of keeping legacy systems alive? What’s the right balance between incremental improvement and “big bang” rewrites? How do you justify refactoring work when it doesn’t align with roadmap goals? What does a “refactoring culture” look like in a truly high-performing engineering organization?