I know a team using it to replace ancient massive mainframe based systems with modern distributed systems and the gist is that the language is fine, but mostly ideal for use cases that leverage the ErlangVM or BEAM stack.
The downside they run into is the ecosystem isnt there, at least a couple guys wish they had just used Kotlin/Java for library interoperability with so much existing code already built and battle tested for specific purposes.
Elixir is not perfect, but for me working alone dependency hell was the bottleneck with Python. Now the bottleneck is adding features, which is right where it should be.
for HL7: https://hexdocs.pm/elixir_hl7/main.html
for MLLP: https://hexdocs.pm/mllp/readme.html
For the less common ones, we tend to just roll our own which in most cases isn't that bad if you have reference implementations.
I think the most under-appreciated aspect of Elixir is how it helps reduce complexity. And there isn't a silver bullet here, but the tooling, immutability, pattern matching, process-based concurrency model, etc are all design decisions that, IMHO lead to simpler, more robust code.
(Caveat: of course, like any language, you can make a mess of things.)
I'm curious what libraries they wish existed.