Some bugs haunt the team for weeks. Recurring, hard-to-track, embarrassing. They ruin the road maps, kill morale and undermine stakeholders' trust. If you ever seen them, you know the nightmare.

Don't despair - I've found a way to deal with them efficiently. A mix of methods and habits used by scientists is the way out of debugging horror. It's a loop of observation, hypotheses and experiments, enhanced with practices used by scholars. It makes debugging more predictable, less frustrating and much faster. I'll share how they helped me in three case studies: fixing a non-deterministic test, stabilising CI infrastructure and resolving a recurring production downtime.

This talk will enhance your programming toolbox for the hardest bugs. Not only this - but you'll see how to scale effective debugging to the entire team. Last but not least - you'll know a set of habits to share with your fellow engineers and help them step up the debugging game.