Preliminary focus:

• Revamped ODBC layer – replace the Erlang ODBC layer with our own ODBC layer which a) is more performant, and b) has support for a wider variety of types
• Improved anonymization – we currently still have anonymization short comings of multiple kinds. On one hand we have an open attack which needs to be addressed, and secondly there are limitations in place which makes working with our system hard. The attack should be addressed in a holistic way, and the limitations unified.