I recently released an open-source framework for machine learning debugging. Proper discipline can be the difference between wild success and a your project slo

I recently released an open-source framework for machine learning debugging. Proper discipline can be the difference between wild success and a your project slowly sinking, like a stone to the bottom of a river. Why? Because a customer adopting your product is risky, dangerous, and uncertain for them. They’re squeezed for time to learn your paradigm. They’re staking their reputation on the line with a boss, coworkers, and company. The great irony is, that by adding features to support one additional customer, you harm the ten customers you already have. Ten happy customers become eleven “okay” customers, then twenty unhappy ones. Make the interface too complex? People will quit. Too many widgets, settings, and knobs will torpedo a project in no time.