It finally happened: tired of mutable data structures and thread synchronisation, we decided to use functional programming and F# for the next generation of our system. Gigabytes of data are going through our applications every hour with high demand for performance, scalability and failure recovery. The fact that our project has for a long time been in production can identify it as success, but has the choice of FP become an essential part of this success? Could we achieve similar development speed and operational stability if we settled for OOP (and C# on .NET platform)? We believe that functional programming provides a set of defaults that can give significant advantages for development with short deadlines and continuous deployment, and we are going to share our experience and lessons learned in this talk.