This can be accomplished with a batch upload to Splunk. Note: this deletes the file upon indexing. If you have a script that copies the csv file to the batch upload directory, then this should work for what you need. If not, use the monitor stanza.
The following configurations are for the Universal Forwarder:
index = [YOUR INDEX]
sourcetype = [YOUR SOURCETYPE]
move_policy = sinkhole
whitelist = .*\.csv
TIMESTAMP_FIELDS = [TIMESTAMP FIELD IN CSV]
INDEXED_EXTRACTIONS = CSV