Known limitations

The limit: The AI scores answers against your rubric and is sometimes wrong — especially on partially-correct answers and unconventional-but-valid reasoning.

What we do about it: The professor always controls the grade: every AI score arrives with the transcript, a written rationale, and a confidence level; you approve, adjust, or override, low-confidence and failed scoring runs are flagged for review instead of auto-finalized, and every override is logged and used to evaluate our scoring.

The limit: Strong accents, noisy rooms, low-quality microphones and rapid speech all degrade the transcript the AI grades from.

What we do about it: We moved to a transcription model specifically for speech quality, the student sees a device check before the exam, and the professor reads the transcript beside every score — a bad transcript is visible, not hidden.

The limit: The AI persona in roleplay scenarios follows your brief but is not a person: it can be more agreeable, more predictable, or less domain-expert than the counterpart it stands in for.

What we do about it: You author the persona, its goals, and how hard it pushes back; a rehearsal lets you hear the character before students do, and the grade comes from your rubric, not from the character's opinion.

The limit: An oral exam inherently requires speaking, and students with speech, hearing or anxiety-related needs may be disadvantaged by the format itself.

What we do about it: Every student-facing session includes a device check, generous time limits you control, and practice attempts; formal accommodations (extended time per student, alternative formats) are not built yet and we do not claim them.

The limit: QPAssess is available in English and Spanish, and that covers what the product writes to you: the interface, its labels, errors, the grades export and the PDF report. It does not cover the assessment itself. Question delivery, the AI examiner, transcription and scoring are built and tested for English; other languages are unsupported and untested today.

What we do about it: We separate the two on purpose, and say so here, rather than letting a Spanish interface imply a Spanish oral exam that would silently degrade. Multilingual assessment is planned and will ship per language only once transcription and scoring quality are verified in that language.

The limit: LMS integration (single sign-on launch, roster sync, grade passback) is built and tested against Canvas only. Blackboard, Moodle, D2L and Google Classroom are not supported today.

What we do about it: QPAssess also works standalone with a share link and a grades export, so a course on another LMS can use it today without integration; further LMSs follow demand.

The limit: Oral assessment means we necessarily record students' voices, and in some modes their camera or screen — data that is personal by nature and subject to institutional policy.

What we do about it: Recordings and transcripts are stored privately, students are told what is recorded before it starts and must consent, and the professor can delete an attempt. We publish our subprocessors and what each one receives.