Historically, the data and cloud story for F# has been strong. Its lightweight syntax and powerful type system provide a succinct way to work with data in a static manner, while .NET has provided the power and flexibility to deal with disparate data sources quickly and easily. F# also provides an excellent story for pivoting between "exploration" and "engineering" styles of development.
With the growing maturity of .NET Core 3 and the move for F# Interactive to .NET Core, scripting processes and tools have started to improve. Users can now directly reference .NET Standard assemblies without workarounds. Work is also ongoing in FSI to allow the direct referencing of NuGet packages, and in the meantime tools such as Paket provide relatively stable automatic generation of "reference load scripts". While not all popular scripting data packages have been ported to .NET Standard, others such as the popular XPlot charting library, are available.
Type providers have historically been a popular and useful tool for working with data in F#, particularly when exploring datasets. However, with the advent of .NET Core, type provider libraries have gone through a difficult transition, with rework required for most of them. Some of the more common providers have been successfully ported across, while others require more work. We are hopeful that now .NET Core has stabilised and that the latest Type Provider template packages are up to speed, other library authors will start the process of porting to .NET Core.
There's lots more changing in the F# data space. To find out which innovations we recommend you adopt and reject, download Compositional IT’s new whitepaper F# In 2020: The Road Ahead.