R in Institutional Research and Advancement: How Harvard, Georgetown, and UPenn Build with Posit

How do you ship one report that shows each school only its own data? How do you kill the weekly Excel-to-PowerPoint scramble? Three higher ed teams show what they built with R and Posit. In Posit's first academic webinar, data teams from Harvard, Georgetown, and the University of Pennsylvania walk through the actual architecture behind their institutional research and advancement work, code included. We close with a first look at Posit Commons, a brand new experimental open source package for building governed, self-service data analysis agents. What you will see: Inger Bergom, Harvard Office of Institutional Research and Analytics: replacing 10+ school-specific PDFs, Excel files, and SharePoint links with a single Shiny app on Posit Connect. Uses the session$groups object plus Grouper-synced group membership to serve each school its own greeting, filtered tables, and downloads from one link. Scaling next to 50 to 60 department-level reports. Lu Hao, Georgetown Office of Assessment and Decision Support: an automated pipeline from BigQuery to Quarto to pins, scheduled on Posit Connect. One cleaned dataset feeds dashboards, scenario simulators, course evaluation reports, and long-form survey reports, all refreshing on their own. Built to survive staff turnover by putting institutional knowledge in code instead of someone's head. Jason Vail, University of Pennsylvania Advancement Analytics: Posit Workbench and Connect on-prem, plus Posit Team on Snowflake Snowpark Container Services. Two AI agents in production: one that generates compliant annual giving appeal copy across schools and centers, and one that summarizes 40 years of fundraiser contact reports. Both served as Python Shiny apps, with data staying inside the Snowflake perimeter. Posit Commons demo with John LaPlante, Posit solutions architect: a framework for data analysis agents in R and Python, built on ellmer, chatlas, and shinychat. You define a semantic layer of validated functions, and every answer comes back tagged verified, cited, or untrusted, so users know what they can stand behind. Version 0.1.0 and highly experimental, announced at posit::conf. Q&A covers making the budget case to senior leadership, where K-12 and district analytics teams fit in, and how to decide between code and no-code for work that decisions get made on. Questions about R and Posit tools at your institution? Reach out to Amy McCorkle on our public sector team at [email protected]. Learn more about Posit: https://posit.co #RStats #InstitutionalResearch #HigherEd #DataScience #Shiny #Quarto