Many people interested in F# programming are also interested in "big data" problems. Here's an interesting job post that shows how these things are coming together in the marketplace:
[ We're ] looking to create a new risk engine using new technology and methods. Building a new dynamic calculation model involving a rich data source, a service manager and a quick index (similar to Google and Amazon). You will be building out a high performance dynamic risk calculation engine, using .net to create the engine with a mix of F# as a functional as a functional paradigm. This will be designed to deliver a low volume search with a high volume data output. This is a highly multithreaded partition data set designed to hold up to a decade's worth of data.
The role would include the drawing of the system, running the through a non relational database and writing a functional F# model. You would be liaising with risk managers who have a strong technical understanding. Preferred technologies would be .net 3.5, functional programming languages: F#/Haskell, exposure to non relational databases: MongoDB/Cassandra and framework tools like MapReduce or Hadoop.