Ransacked uses Wwise to handle the audio. There are many instances where the audio events are triggered but the outcome is dependent on the specifics of an entity, which unit it is for example.
I came up with the solution of what I call a soundbank component. It's basically just a dictionary of audio event keys which can be mapped to Wwise events.
The audio system can post these events to Wwise using the soundbanks or directly from the Wwise event name. It is also capable of posting events from one entity's soundbank through another entity. For example, a unit's weapon might have a sound for when its projectile hits another entity.
It's a pretty simple system but very organized. I like the implementaion quite a bit.