SQLCollections.jl is finally released in General 
A nice synergy – that basically comes for free – is querying tabular files like CSV or parquet. See the QuackIO.jl package: it could already use DuckDB to performantly read such files into Julia, like read_csv(StructArray, "my_file.csv").
Now, with SQLCollections.jl, we can query these files without fully loading into memory, while still using the same Julia syntax:
using QuackIO, SQLCollections
data = read_csv(SQLCollection, "my_file.csv")
# data is an SQLCollection and can easily be queried :)