Course feedback that feels like a conversation.
QAAI turns open, typed or spoken student feedback into structured, anonymised quality assurance insight — without inventing ratings or skipping the questions that matter.
Runs on the university's own server. Local AI. No third-party processing.
Database Management Systems
CSC2104 · Semester II
For students
Answer in your own words, by typing or voice. One question at a time, with a clear review before anything is submitted.
For QA reviewers
See ratings, recurring themes, representative comments, and the original anonymous response behind every number.
For institutions
A transparent evaluation trail: fixed criteria, deterministic coverage, auditable records, and demo data kept separate.
A guided conversation, a disciplined evaluation.
The AI handles wording and extraction. A deterministic controller handles structure: it knows which criteria are covered, what still needs a rating, and when the evaluation is ready to submit.
Start anonymously
No student account required. Each evaluation uses an anonymous session token, never a student identity.
Talk naturally
The assistant covers every criterion in order. One answer can raise several issues; the controller keeps track so nothing is skipped.
Review & submit
Check exactly what was captured, add a rating only where the template requires one, and submit once.
Turn talk into insight
Reviewers get database-backed aggregates, criteria analysis, recurring themes, and the original anonymous records.
Uganda Martyrs University, Faculty of Science
The demo uses a real course structure with clearly illustrative evaluation records. It is a working prototype, not an official institutional deployment.
- Course unit
- Database Management Systems
- Course code
- CSC2104
- Lecturer
- Demonstration Lecturer
- Evaluation areas
- 8 criteria
Demo records are tagged is_demo in the database and can be filtered out of live quality assurance reporting.
Evaluation template
Illustrative data- Teaching clarity
- Lecturer preparedness
- Learning materials
- Practical learning
- Student engagement
- Lecturer accessibility
- Assessment feedback
- Overall experience
Only the overall experience asks for a numeric rating. Everything else is captured from what the student actually says.
Anonymous by design
No student account, name, or registration number is collected. Reviewers see anonymous response records only.
No invented ratings
Ratings come only from explicit student input. The AI can summarise wording, but it cannot manufacture a score.
Demo and live stay separate
Every session is tagged as demo or live, so demonstration records never contaminate real QA analysis.
Try the full loop: conversation, review, dashboard.
Complete an anonymous demonstration evaluation, then sign in as a reviewer to see how the same responses become structured QA insight.