External data lands in wrong tables or gets lost
Your team is responsible for pulling employee data nightly from an external HR system into ServiceNow's User table. The current approach — a custom script that calls the Table API directly and writes records one-by-one — is fragile: fields occasionally land in the wrong columns, there's no place to inspect raw payloads before they hit production tables, and failed records disappear silently. The integration owner wants a more structured pattern that stages incoming data, allows field-level inspection and correction before committing, and supports custom transformation logic for edge cases. What architectural approach would you recommend before the next nightly sync window?