Nice stuff, thanks! Regardless of whether you go for the 'big-bang rebuild' or the slow and steady migration, you'll need to get engineers on the case at the cost of shipping new features. Most companies are at least half-decent at drawing a line between new features and revenue (or whatever their main measure of value is). How would you draw a line between paying back technical debt and value? In my experience, it's often hard to get people who don't truly understand technical debt on board with projects like this.