Hello @here,
Greetings!!
I am currently using the Zingg Community Version for customer entity resolution on Databricks. I recently demonstrated the implementation to the client, and they were satisfied with the results.
However, we observed that Zingg uses the DBFS location by default to store intermediate results, such as marked and unmarked data, as well as the matched output. Since DBFS is being officially deprecated, the client has raised concerns about continuing with this approach.
They would prefer to use a Unity Catalog Volume path instead of DBFS.
Could you please confirm whether this is supported in the Zingg Community Version? Specifically:
Can we configure Zingg to use a Volume path for storing intermediate data?
Can the matched output be written directly to a Unity Catalog Volume instead of DBFS?
Sonal G., could you please guide us on this?
Once we have clarity on this aspect, the client will be comfortable proceeding with taking the Zingg implementation to production.
