CASE AI Coffee Chat: Practical Uses of AI, from Due Diligence to Automated Reporting-RCSI Presentation

RCSI Presentation

Pdf Summary

Charlotte O’Mahony, Data & Insights Analyst at RCSI, presented a case study on using AI as a coding partner to automate alumni engagement metrics reporting. The project addressed a repetitive manual reporting task for CASE Alumni Engagement Metrics (AEM), where criteria were clear, results were measurable against prior-year data, and Python-based tools were available.<br /><br />She outlined effective AI prompting principles: define the task, provide context, include examples, specify format, set constraints, and assign a role or persona. The workflow began with ChatGPT generating a first draft of code for volunteering activity, which was then adapted for experiential, communications, and philanthropic activity. The code was tested in MS Fabric against real data, then refined through iterative debugging with AI support.<br /><br />The project required around 25 hours over three weeks, involving 15–20 prompts and 8 code iterations. Key lessons included the importance of planning and documenting technical requirements, verifying outputs rather than trusting them automatically, and carefully considering ethics and data protection—especially avoiding use of real customer data in generic LLMs.<br /><br />The benefits of this approach were significant: it reduced time spent on benchmark reports, lowered the risk of human error, and helped bridge the gap between business understanding and technical implementation. Looking ahead, the model could be reused for other RCSI purposes, such as engagement scoring, prospecting, and longitudinal analysis.

Keywords

AI coding partner

alumni engagement metrics

CASE AEM reporting

ChatGPT prompting

Python automation

MS Fabric

iterative debugging

data protection

benchmark reporting

engagement scoring