By looking at your code, I can suggest a few changes that you can try out.

Instead of spark.readStream, you can use bronze_df = dlt.read_stream("Transactions_Bronze").

Revised Code:

bronze_df = dlt.read_stream("Transactions_Bronze")
dlt.apply_changes(
source=bronze_df,
target="Transactions_Silver",
keys=["Company", "VoucherType", "AccountingYear", "VoucherNo", "RowNo"],
stored_as_scd_type="2",
sequence_by="Inserted"
)

Additionally, you have created a column called SDA_Inserted in the bronze table, but in the silver table, the sequence_by parameter is based on the Inserted column. Make sure to correct the column name if there is no Inserted column already present in the data.