Just free associating here, actually. The single session of 75 minutes is awfully short to actually have you do anything and then have the students learn anything. So....
First, flip the classroom for that session even if you don't generally. Give them a week (more or less???) to study Kotlin on their own, so that they have some sense about its syntax and main features. You may have to give them an outline of what to focus on.
For the single session, have them build a stateless server for some simple app using Kotlin in small groups, even pairs. Making it stateless lets you more easily stress the FP aspects (I think, but haven't actually tried it). Note: still thinking about the stateless part. Think about sending them something the evening before the session to get them thinking about what is coming. Not a spec, though, since some will try to do too much beforehand.
One possibility that will help speed up the day's work is to send them a skeleton application the night before. The task of the class will be to extend the skeleton to something useful. I would probably do that if it were my class, actually.
Use the last quarter hour for a retrospective on what they learned and point them toward the future.
I don't know if this would be successful. It might fail. It might require modification for a future use. Owen Astrachan at Duke as advocated that some small percentage of what you do in a course should be experimental so that you can learn more about what works and what doesn't. But be prepared for it to fail so that students can learn even from failure.