How do you secure infrastructure at scale while maintaining autonomy for engineering teams at Amazon Services?
