As someone who writes lots of code that wrangles messy raw data inputs via DataFrames, the talk on Pandera type checking for DataFrames was very interesting. Dunno if it will work in practice for my use case, but I’ll definitely be checking it out.