Hi @menotron 

Yes, as I mentioned "1>SQL function context don't have SparkSession. I am forced to use normal python code instead of pandas udf as sparksession is not available within sql function. " I have used the Python UDF for Databricks SQL and it is not performing well as I can user pandas udf there to improve the performance. It don't have sparksession.

Also I dont want to use KV or databricks backed secret scope as the user who need access to view also need read access to the secret scope. In that case that user can get the decryption key and share with other user. This is security issue.

Many thanks