Explore capabilities for Sense
Build, configure, and deploy AI agents connected to your LMS — exam markers, course summarisers, AI tutors, and more. No AI infrastructure required. Bring your own OpenAI key.
- Core Platform
AI Agent Builder
A no-code platform where lecturers and academic staff build, configure, and manage AI agents without writing any code.
- Guided wizard — choose agent type, upload materials, set rubrics
- Dry-run and test before going live
- Activate, deactivate, or duplicate agents at any time
- Automation
Exam Marking Workflow
Students submit through the LMS as normal. Sense routes submissions to the marking agent, evaluates against the rubric, and holds results for staff review before release.
- Rubric-based AI evaluation with human review before students see results
- Reviewers can approve, edit, or override AI marks
- Full audit trail for moderation, appeals, and accreditation
- Student Experience
Revision Tutors & Student Q&A
Give every student 24/7 AI tutoring scoped to your course content — not the open internet.
- Responds from uploaded course materials, not generic internet data
- Pre-seen exam tutors can be time-gated and topic-scoped
- Runs inside the LMS — no separate app or login needed
- Knowledge Management
Per-Agent Knowledge Bases
Each agent operates from its own knowledge base — lecture notes, rubrics, and course documents scoped precisely to its purpose.
- Upload lecture notes, rubrics, and reading materials per agent
- Marker and tutor for the same module use separate knowledge bases
- Updates take effect immediately — no redeployment needed
- LMS Integration
LMS Agent Executor
A plugin that runs directly inside your LMS, detecting student events and routing them to the right agent workflow automatically.
- Triggers on submissions, revision sessions, and Q&A requests
- Runs on your institution's own OpenAI API key
- Student data stays on your server — nothing leaves your environment
- Templates
Agent Type Catalogue
Pre-built agent types for education — exam markers, revision tutors, mock exam engines, FAQ bots, and fully custom agents.
- Purpose-built for educational workflows
- Pick a type, configure, and deploy — no starting from scratch
- Custom agent type for unique institutional workflows
- Governance
Approval Queues & Audit Trail
AI outputs with academic consequences go through a staff review queue before reaching students, with a full timestamped audit trail.
- Configurable approval queue per agent
- Reviewers approve, edit, or reject AI output
- Built for accreditation reviews and appeals
- Visibility
Cost Dashboards & Usage Analytics
Track token costs and usage across every agent and module, with full isolation between institutions.
- Cost breakdown by agent and module
- Usage analytics across all active agents
- Institution workspaces fully isolated
Frequently asked questions
What AI models does Kampus Sense support?
Any OpenAI model accessible via your API key — including GPT-4o, GPT-4o mini, and future models as they become available. You choose the model per agent, allowing cost and performance trade-offs per use case.
Does Kampus Sense store student data or exam submissions?
No. All production AI execution happens inside your Wave plugin. Student submissions, course files, and answer papers never leave your institution's server. Kampus Sense only stores metadata — agent configurations, run statuses, and token usage logs.
Do I need technical skills to build agents?
No. Agents are configured through a guided, no-code wizard. Your IT team handles the one-time Wave plugin installation. After that, lecturers configure and manage agents entirely through the browser interface.
How is Kampus Sense pricing structured?
Kampus Sense charges a platform fee based on your subscription plan. OpenAI API costs are entirely separate — you pay OpenAI directly, with full visibility via the built-in cost and token usage dashboard.
Can I have multiple AI agents for the same course?
Yes. Create as many agents of different types as you need for the same course — for example, a Pre-Seen Tutor and an Exam Grader running independently for the same cohort.
Is the marking fully automated or does a human review it?
Both modes are supported. You can require lecturer approval before results are visible to students, or enable fully automated release — configurable per agent. Low-confidence submissions are always flagged for optional manual review.
See Kampus Sense in action
Add AI agents to your institution without replacing your LMS or exposing student data.
