Skip to main content

Institutional field notes

Questions higher education should answer before the technology does.

These are direct, evidence-grounded answers for institutions deciding how AI should support students. They draw on Project Ardi, a 44-day planning pilot at CU Boulder, and on the operating questions the pilot left behind.

Published September 24, 2026 · Ardvarq, Inc.

Seven field notes
  1. 01What do students actually want from academic planning?Student success, advising, and academic affairs
  2. 02Why do university resources remain undiscovered?Student success, communications, and digital experience
  3. 03What should AI do in academic advising?Advising leaders, provosts, and student success teams
  4. 04Can universities personalize support without predictive risk scoring?Privacy, procurement, IT, advising, and student affairs
  5. 05How should a university design an AI pilot?Presidents, provosts, CIOs, and innovation leaders
  6. 06How should universities evaluate an AI student-support pilot?Institutional research, assessment, and product teams
  7. 07What would an AI-native university resource layer look like?Presidents, provosts, CIOs, and digital strategy leaders

Start with your question

The same decision looks different from every seat.

President or provost

Advising or student success leader

CIO, privacy, or procurement leader

Institutional researcher

Evidence standard

Useful enough to act on. Honest enough to question.

We distinguish observed pilot behavior from institutional recommendations. Project Ardi measured product use and conversations, not learning, grades, persistence, or causal impact. The public pilot numbers are a 44-day run, a company-reported estimate of more than 2,000 student users, and 903 tracked conversations. Where the evidence stops, we say so.

Read the research notes