CampusAI · Documentation

Product documentation

How CampusAI works and how to use it — for students, career-services teams, and university administrators. Covers features, getting started, troubleshooting, and support.

University Intelligence Platform · Version 1.0 · Updated July 2026

Overview

CampusAI is an AI-powered campus platform for university students, delivered as a multi-tenant web application. Each partner university receives its own branded portal on a dedicated subdomain (for example umd.campusai.app), its own isolated identity pool, and a configurable AI assistant.

The platform serves three audiences through separate experiences:

  • Student portal — university students, at {university}.campusai.app
  • OCS portal — career-services staff and directors, at /ocs on the university's portal
  • Platform admin console — CampusAI operations staff, on a dedicated admin host

AI features are powered by Google Gemini models. Student data is partitioned per university tenant, FERPA-scoped records require explicit student consent, and service-level access is audit-logged.

Features

Feature availability is configured per university; a campus may launch with a subset of these surfaces.

Student portal

Today dashboard

A personalized home page with an AI daily briefing, today's classes, recommended events and jobs, deadlines, and a week-ahead view.

AI assistant

A conversational assistant on every page — a context-aware advisor panel, a quick ask bar, and full chat with history. Searches courses, professors, seats, jobs, events, dining and more, with web grounding.

Career advisor

Job and internship matching with plain-language fit reasons, an application tracker, an ATS resume coach (suggestions only), skills-gap analysis, career-path generation, voice mock interviews, and advisor booking.

Academic advisor

Course explorer with AI insights and open-seat data, a professor directory, a four-year planner with AI recommendations, a degree-progress audit, and a syllabus-based class tracker.

Campus life

Personalized events, dining with menus and dietary filters, and — where enabled — shuttle and transit, a student-org directory, athletics, campus locations, IT helpdesk, and a wellness center that routes crises to professional resources.

Profile & settings

Onboarding wizard, profile-strength meter, Canvas linking, a privacy summary, consent records, and self-service data export and account deletion.

OCS portal (career-services teams)

A director-level dashboard of cohort health, readiness KPIs and predictive models (retention, graduation, placement); a student directory with engagement signals and at-risk outreach; weekly insight briefs; cohort resume/skills quality against employer demand; jobs-market gap analysis; outcome tracking; employer relationship management; and a methodology page for every metric. Analytics run on de-identified aggregates with small-cohort suppression.

Data integrations

Managed connectors keep campus data fresh: Canvas LMS (student-authorized), Handshake and 12twenty, event calendars, course catalogs with real-time seat data, professor reviews, dining feeds, Google Places, transit, athletics, and student-organization directories. Sources refresh on cadences from every 15 minutes to weekly.

Getting started

Signing in (students)

  1. Go to your university's CampusAI portal, e.g. https://umd.campusai.app.
  2. Click Sign in with your university email and complete Google sign-in with your university account. Only addresses on your university's domain are accepted.
  3. Some accounts (evaluation or staff) use the Sign in with email & password option instead.
  4. On first sign-in, complete the onboarding wizard (about you, interests, academics, career direction, campus life). Progress saves per step, and you can re-run onboarding later from your Profile.

Sessions refresh automatically; if your session expires you're returned to the sign-in page.

Using the AI assistant

Type into the ask bar at the bottom of any page, use the quick-action chips (they change with the page you're on), or open the full chat for longer conversations. Use New chat to start a fresh topic; past conversations live in the history drawer. The assistant only accesses features enabled for your campus, and staff-analytics capabilities are role-gated.

Key student workflows

  • Get job matches: complete onboarding (target roles required) → Career → review ranked matches and the "why this fits" reasons → track in the Applications view.
  • Improve your resume: Career → Resume → upload your resume → review the ATS check, score, and suggestions.
  • Practice interviews: Career → Interview Prep → optionally pick a target job → complete a voice mock interview → review the coaching wrap-up. Audio is not stored; transcripts are encrypted.
  • Plan your degree: Academics → Planning → build your four-year plan → review AI semester recommendations and the skills each course builds.
  • Track classes: Academics → Class Tracker → upload a syllabus to auto-extract assignments and due dates, or link Canvas in Settings.
  • Manage your data: Settings → privacy summary, consent record, data export, or account deletion.

Staff & administrator workflows

  • Career-services staff sign in with their university account and open /ocs. Start at the Overview ("what needs you this week"), drill into Students for at-risk outreach, and use Methodology to understand each metric's calculation.
  • University administrators request configuration changes (branding, enabled features, identity providers, new integrations) through CampusAI support; changes are applied by CampusAI operations via the admin console.

University onboarding (new customers)

CampusAI provisions each new university: a dedicated identity pool with your SSO (Google sign-in scoped to your email domain, with optional MFA and password-login policies), a branded subdomain, your logo/colors/assistant name, the feature set covered by your contract, and data connectors for your campus systems. Initial content is seeded during onboarding, after which scheduled pipelines keep it current.

Troubleshooting

Message / problemWhat it means and what to do
"This email address isn't authorized for this campus."You signed in with a Google account outside your university's domain. Choose "Sign in with a different Google account" and use your university email.
"Single sign-on is not configured for this university yet."Your campus's SSO is still being set up. Contact your university administrator.
Sign-in failed after Google sign-inClick Try again. If it persists, contact your university IT support.
"Incorrect email or password."Password login is only for accounts created with a password. Students should use university-email sign-in.
"Tenant Unavailable"The portal couldn't load your university's configuration. Check the URL (e.g. umd.campusai.app), retry, or contact your administrator.
"Operations access denied" (403)You signed into the operations console with a student account. Use your university's portal instead.
A page or feature is missing / "…is not enabled"Features are enabled per campus. Contact your administrator if you believe this is an error.
"This section hit a snag"A view failed to load; your account and data are fine. Click Try again — the rest of the app keeps working.
AI assistant shows "Reconnecting" or "Error — try again"A temporary connection issue. Wait a moment and resend; your conversation history is preserved.
AI responses degrade to simpler contentUnder heavy load, AI enrichments fall back to non-AI content rather than failing. Retry later for the full experience.
HTTP 429 / "too many requests"Per-user rate limits were reached (see Service limits). Wait for the hour to roll over; sign-in and consent flows are never rate-limited.
Redirected to the home page unexpectedlyThe address was mistyped or points to a page your role can't access. Navigate from the sidebar.
Platform-wide issuesCheck the CampusAI status page for current availability and incident updates.

Service limits

Per-user API request limits by subscription tier: Starter 100 requests/hour, Professional 500 requests/hour, Enterprise 2,000 requests/hour, with short burst protection between repeated identical requests. Health, consent, and privacy endpoints are exempt, so sign-in and FERPA consent are never blocked. Voice interview sessions and AI generation calls have internal timeouts that degrade gracefully to non-AI content.

Privacy & security

  • Tenant isolation. Every record is partitioned by university tenant, and services re-verify tenant ownership on each read. Cross-tenant access is impossible by construction; internal service-to-service calls carry signed tokens.
  • FERPA. Educational records require explicit student opt-in consent, tracked by scope. Staff access to an individual student's records goes through a dedicated FERPA disclosure workflow; staff dashboards otherwise use de-identified aggregates with small-cohort suppression.
  • Audit logging. Authenticated API access is written to an append-only, tenant-scoped audit log recording endpoint, caller, outcome, and request ID.
  • Student data rights. Students can view a privacy summary of exactly which sources are tracked, export their data, and delete their account self-service.
  • AI data handling. Voice interview audio is never stored (transcripts are encrypted at rest). Student chat, profile, and memory data are excluded from shared search indexes. Crisis-related conversations are always routed to professional resources.

For the full security overview, tenant-isolation architecture, and encryption standards, see the CampusAI Security page.

Support

Need help or want to report an issue? Contact support@campusai.app. For current platform availability and incident history, see the status page.