The words on this site, defined.

AI that completes multi-step work on a team's behalf rather than only answering questions in a conversation.

An agent can build a student group, draft an automation, or generate a report. Ask of any agent: what is its model trained on, and who keeps the decision? Civitas agents act on institution-specific predictions, staff set the guardrails, and every action is logged and attributed.

Civitas Learning AI →

Handling a student's advising alert, aid hold, and support referral as one case with one owner rather than three separate contacts.

Coordination fails when three offices each hold part of the story and none can see the rest — which is how students get contradictory instructions or a flood of messages.

Coordinate student care →

Civitas Learning's five-part operating model for student success: Connect, Learn, Operationalize, Strengthen, Evaluate.

Connect data and teams around shared context. Learn from institution-specific patterns, not generic benchmarks. Operationalize insight in daily workflows. Strengthen human judgment with transparent AI. Evaluate impact and close the loop. It describes how an institution designs its systems so improvement compounds instead of resetting each year.

The CLOSE model in full →

Predicting how many seats each course and section will need in a future term, from degree plans and student progression.

Forecasting builds the schedule from what students need to progress, not last year's schedule. It prevents both costly failures: the section that does not fill, and the missing seat that blocks a student thirty credits from a degree.

For academic leaders →

A notification that a student may be at risk, raised by a rule, an instructor, or a predictive model.

What matters is what raised it. A rule fires on a threshold someone chose, such as a grade below C. A model weighs grades, credits, course patterns, aid, and engagement together and ranks students by risk. Alerts also need follow-up: in measured partner data, an early alert with no follow-up produced no significant lift.

Predictive analytics →

A predictive model trained on one institution's own student records rather than on sector averages or a vendor's pooled customer data.

The model learns what predicts outcomes at that institution, such as the 14-credit part-timer or the transfer with a math gap, and is retrained each term. Civitas builds one model per outcome per institution — persistence, graduation, engagement, thriving, grades, and career outcomes — and predicts student outcomes with 89 percent accuracy. No institution's data is pooled into a benchmark sold to another school.

Actionable analytics →

The practice of measuring what every institutional initiative changed, across every office, on one dataset.

Institutional Impact Management extends student success beyond advising. The same unified data, institution-specific models, and comparison-group measurement answer enrollment, engagement, persistence, completion, career outcome, course demand, and program portfolio questions. The test: can an office state the lift its work produced, not just the activity it performed?

The Institutional Impact Platform →

Civitas Learning's platform for unifying institutional data, predicting outcomes, acting in existing workflows, and measuring the result.

The platform reads from more than 100 systems, builds one governed record of the institution, trains institution-specific models on it, and measures each initiative against a matched comparison group. It also carries the tools people work in: an advising CRM with case management, notes, alerts, and outreach; degree planning; registration; class scheduling; and a native student mobile app. Use it as the advising system of record, or keep your CRM and let Civitas write back to it. It was previously named the Student Impact Platform, and older material still uses that name.

Platform overview →

The difference in an outcome between students who received an initiative and a matched comparison group who did not.

Lift is stated in percentage points or percent, with population size and a confidence interval. A lift whose confidence interval crosses zero is not a result, and reporting it as one is the most common way vendors overstate a product. Across more than 220 measured initiatives, 40 to 60 percent produced a measurable lift.

Efficacy and ROI →

Tracking the whole student body and alumni population over years, rather than surveying graduates once.

A first-destination survey captures one moment from a self-selected sample, which is why its numbers usually look better than the cohort did. Longitudinal mapping follows the population over time, so you can see whether graduates stay above a threshold long after commencement.

Career outcomes →

A group of an institution's own students who did not receive an initiative, matched to the students who did on the characteristics that predict the outcome.

The comparison group turns a change into evidence. Without it, a rise in persistence could come from the initiative, a different incoming class, or an unrelated policy change. Civitas builds the group from the institution's own records and reports the difference with a confidence interval.

How efficacy is measured →

Whether a student continues their enrollment from one term to the next, at any institution or at the same one.

Persistence is measured term to term, making it the fastest honest signal an institution has. Retention is the annual, institution-level version, and the figure IPEDS collects. An institution can lose persistence in spring and not see it in the retention rate until fall.

Advising and student success →

Holding an academic program to a standard based on what its graduates earn after they leave.

The Department of Education's earnings accountability framework was finalized in 2026 and takes effect in July 2027. It compares program graduates' earnings to a benchmark, and programs that fall short can lose federal student aid. Institutions can measure their programs against the threshold years before a federal determination.

Career outcomes and accountability →

A designation that lets an institution share student records with a vendor performing an institutional service, under the institution's direction and control.

Civitas Learning is a school official designated by the institution, under a data sharing agreement the institution signs. The institution owns its data, sets the access model, and can audit it. Student data is never pooled into a benchmark sold to another institution.

Security and privacy →

The distance between the effort an institution puts into student success and the outcomes it can actually show.

The gap opens when real work is fragmented: teams act in silos, insight arrives too late to use, and nobody can say which effort moved a number. It is not a gap in care or effort. It closes when data, workflow, and measurement connect, so improvement compounds instead of resetting each year.

How to close it →

Civitas Learning's annual analysis of which student success initiatives produced measurable outcomes across partner institutions.

Each edition reports the persistence effect of practices such as advising, outreach, academic planning, career engagement, and co-curricular participation, measured against comparison groups and broken out by population. It is the source for the finding that only 40 to 60 percent of initiatives produce a measurable outcome.

Resource library →

One governed dataset that joins the student record across every system the institution runs, with shared definitions.

Without it, the registrar, advising, and institutional research answer the same question differently, each reading a different copy of the data with a different definition. The layer makes a retention figure reconcile across the board deck, the accreditation self-study, and IPEDS.

The Data Lakehouse →