Program Title: AI-Assisted Code: Automating Merge Record Cleanup as a Non-Programmer
Track: Sierra, Lightning Talk
Program Description: n this lightning talk, I will describe how I used AI tools to help me streamline a time-consuming process. Every week, Gleeson Library downloads weekly merged records from OCLC WorldShare Collection Manager. The merge files are created when two or more existing records are merged with a new OCLC control number. The problem is that when loading the merged records into Sierra, Sierra cannot match them to more than one existing record at a time. This left our catalog littered with old records that should have been overlaid by the merge records, but were not. These had to be tracked down individually and deleted. To solve the problem, I used Claude to help me write a Python script that would pull the merged OCLC control numbers from the merge file and plug them into a SQL query so that I could quickly pull all the old records to be deleted rather than searching for them one by one. The original task that took hours of staff time now only takes a few minutes to complete. This presentation will discuss my process for writing and testing the script as a Python beginner and share tips and tricks for using AI to assist in streamlining workflows.
Speaker/ Information: Amanda Raver. Metadata Librarian, Gleeson Library, University of San Francisco