What educators are building with Axiom
Axiom is agentic AI, so instructors aren't limited to a single chatbot. Each course can have its own network of agents: narrow, grounded in your materials, and designed around your learning outcomes. Here are some patterns already running in courses.
Tested in live courses since Winter 2026
- courses taught with Axiom
- 20+
- institutions
- 5
- students
- 9,000+
- of AI-supported learning per student, on average
- ~12 hrs
- funded research studies in Ontario
- 3
Largest institutional pilot: A Canadian research university · 9 courses · 3,000 students · 1 research study
One assistant per job, not one bot for everything.
Instead of a general chatbot, students get a set of focused assistants, each grounded in one slice of the course:
- ✓
Course info navigator. Answers questions about the syllabus, schedule and policies.
- ✓
Assignment helpers. Guide students through a task without doing it for them.
- ✓
Exam prep helpers. Practice and explanation grounded in what the exam covers.
- ✓
Unit assistants. One per unit, drawing only on that unit's materials.
Because each assistant is narrow, its answers stay on course, and instructors can see which parts of the course students need help with.
The conversation is the assignment.
Students interview an AI patient who presents with symptoms and responds realistically. They gather a history, prioritize, and justify their clinical decisions.
- ✓
Every question and decision is captured in a full transcript.
- ✓
An evaluation agent scores the clinical reasoning against the instructor's rubric.
- ✓
The instructor reviews the reasoning, not just a final answer.
Students practise clinical judgment safely, and instructors see how they got there.
Several AI stakeholders, one decision.
Students negotiate with multiple AI characters who want different things: a CFO guarding the budget, a board member pushing for growth, a supplier defending their terms. The student has to balance them and make a recommendation.
- ✓
Each agent plays its role consistently, under its own name.
- ✓
The rubric rewards applying course frameworks, so strong scores can't be gamed with generic answers.
- ✓
The instructor sees the full transcript and the scores.
See understanding, and how AI was used to get there.
Axiom includes agents that work around student activity, not just inside it:
- ✓
Course builder. Guides instructors in designing their course's agents.
- ✓
Mastery check. Probes a student's understanding before they move on.
- ✓
Process analysis. Examines how students used AI in their work.
The result is instructor insight into both learning and AI use, course by course.