Massive stars are chemical factories producing key elements, they are progenitors of supernovae, neutron stars and black holes, and they play a crucial role in the formation and evolution of galaxies. Given their prevalence in binary systems, at the end of their lives they may produce double-compact objects, which are potential gravitational-wave sources. During their live cycles, interactions with their companion stars can drastically alter the evolution of both stars. Yet, the complex interaction physics as well as the outcome of the interactions remain poorly understood. One way of constraining those is by observing post-interaction binaries.

A century-old question in the context of massive stars addresses the Be phenomenon, which occurs in ~20% of the early-type stars. Observationally, classical Be stars are defined as B-type stars with Balmer line emission, indicative of a circumstellar disk, which strongly correlates with rapid rotation of the star. While the processes that lead to such high rotation rates are still widely debated, classical Be stars were proposed to be mass gainers in previous binary interactions. If true, that would make them post-interaction binaries with stripped-star or compact-object companions.

In my talk, I will discuss the different channels proposed for the formation of classical Be stars, with a particular focus on the binary channel. I will present observational evidence suggesting that the binary channel is indeed predominant in the formation of massive Be stars, and will show that the few known Be binaries are exotic systems with stripped or compact companions. I will furthermore discuss what those systems can teach us about binary interaction physics and thus about massive-star evolution in general.