Hi @Chibberto, Yes, connecting to a delta table within the MLflow sync serving endpoint is possible.
However, as you correctly pointed out, the serving endpoint environment cannot access Spark. Therefore, you would need to instantiate a Spark session, which would have an amount of overhead.
However, you can use the Delta Standalone library, a single-node Java library that can read from and write to Delta tables. This library provides APIs to interact with table metadata in the transaction log, implementing the Delta Transaction Log Protocol to achieve the transactional guarantees of the Delta format.