We are planning to customise code on Databricks to call Salesforce bulk API 2.0 to load data from databricks delta table to Salesforce.

My question is : All the exception handling, retries and all around Bulk API can be coded explicitly in Data bricks? That won't be an issue.