This project demonstrates how artificial intelligence empowered a small advancement services team to overcome technical barriers and revolutionize the graduation import process at the University of Canterbury. Lacking the budget for a developer, our team leveraged AI tools to generate and refine Python scripts, allowing us to semi-automate and optimize the import process—reducing the timeline from three weeks to three days. The challenge was not just the process but the data itself. The quarterly graduation import contained inconsistent, fragmented address data, with critical elements (e.g., city, state, and postal code) misplaced across columns. Using AI-assisted coding, we developed a solution that automated much of the data cleaning, ensuring customer relationship management compatibility while improving usability for reporting and mailings. Key enhancements included restructuring misplaced city and state values, introducing standard fields like "Region/Province," and preserving original data for traceability. By enabling staff with minimal coding experience to create a scalable solution, this project extends beyond graduation imports. The same AI-driven approach was later applied to financial charge coding, demonstrating its adaptability; the framework continues to provide institutional value.
From the Judges
We liked this entry’s clever use of evolving AI technology to bridge knowledge gaps and enable non-technical staff to program data cleaning tools they would typically need a developer to create. It demonstrated an effective use of resources, saved significant time, and is replicable at other organizations and across many types of data projects.
