Assessment and rubric context
Assignment instructions, rubric dimensions, focus areas, and question context may be processed to generate probe plans and grading suggestions.
AI PROCESSING
Clear boundaries for student data
Pruuva uses AI to generate suggestions, probes, transcripts, and evidence summaries. Those outputs support instructor review; they are not final grades.
Student artifacts
Submissions, responses, code execution evidence, and extracted text may be processed to create artifact-specific questions and evidence reports.
Probe transcript and session context
Transcript turns, response timing, configured evidence mode, and session events may be used to summarize demonstrated understanding.
Configured AI providers
Pruuva can route AI work through approved providers based on product configuration, institutional requirements, and available capabilities.
COMMITMENTS
What institutions and students should know before assessments begin
No model training by default
Student submissions and assessment conversations are not used by Pruuva to train generative AI models.
Consent before recording
Students see the configured evidence mode, monitoring notice, provider implications, and retention policy before starting.
Teacher authority
AI outputs are evidence-backed suggestions that teachers can accept, adjust, or reject.