For university academics

Good workhas company.

For the work behind academic work.

PaperGrade

Available

Your criteria.
Your assessment.

Prepare, calibrate, and review exam answers. Keep the reasoning in view.

Open PaperGrade

A score proposal is not a final grade.

Illustrative exam example

Select a criterion to follow the proposal.

  1. Prepare the exam and rubric
  2. Calibrate with reference answers
  3. AI score proposals
  4. Instructor review and decision

Exam question and student answer

Does this study show that the app works?

Students chose whether to use the app. App users scored higher, but motivation may explain the difference. This shows an association, not proof of causation.

Rubric criterion

Criterion proposal: 2 of 2 points

Recognises self-selection rather than random assignment.

You accept, adjust, or reject this proposal.

Librarian

In development

Research that
stays connected.

Start with your work. Connect the sources you add to what you are making.

Sources you added

Your work Research paper

Shared method

A method you can compare.

A book chapter that uses the same coding method for seminar discussions.

“Have two readers code each discussion independently.”

Illustrative example

Fictional research map

A connection is useful when you can explain it.

About Librarian

Built around your manuscript, research project, or other work. You add articles, books, reports, and documents; Librarian is being developed to explain their connections to that work. Finding new sources automatically is a future possibility, not part of the current plan. No public application is available yet.

Credit transfer

Research project

Learning
should travel.

Keep institutional rules fixed. Explore course equivalence that adapts to faculty judgment.

Course evidence map

Illustrative course comparison.

Fixed institutional requirements

  • Minimum ECTS ratio
  • Minimum grade
  • Prerequisites
  • Course level

Fixed rules. Feedback never changes them.

Source course

Research Methods

Target course

Applied Research

Study design
Describing data
Ethics not documented

Source evidence

“Develop a research question, hypothesis, and study plan.”

The source syllabus documents the required design work.

Faculty feedback

“Matching topics alone is not enough here. Students must interpret results, not just calculate them.”

Guidance for later comparisons

For this department, distinguish calculation from interpretation when weighing content differences. Keep this example for later judgments.

The proposal updates judgment guidance, not institutional rules. Check rule preservation before using an update. Recommend accept, reject, or modify; the faculty decides.

A comparison to inspect, not a recognition decision.

About the research

A proposed multi-agent system reads course documents, retrieves candidate courses, and assesses equivalence with evidence. Faculty corrections would refine departmental guidance and examples without retraining the model or relaxing institutional constraints. The paper presents a conceptual design, not measured performance or a public application.

Where minds
meet peers.

A useful second perspective.
Not a substitute for yours.