OASIS: Open Assessment and Scoring Infrastructure Stack

Rubric-based multimodal assessment using large language models

Project information and technical report for OASIS

An AI oasis in the desert of manual grading.

OASIS grades video, audio, and text against evaluator-defined rubrics, using hosted or self-hosted large language models. Modality-aware execution, provenance capture, and human review keep scores inspectable from rubric definition through adjudication.

Technical report

The report covers the platform’s architecture, rubric-to-prompt compilation, multimodal grading pipeline, command-line and agent interfaces, provenance model, and operational experience.

• Read the technical report online
• Download the technical report (PDF)

Related projects and demonstrations

MAPLES is the platform’s grading and review application; Wayfinder is the assistant embedded in it.

• Explore UT-REAL, a multi-site MAPLES project
• Watch Ameer Hamza Shakur demonstrate the MAPLES grading and faculty-review workflow
• Watch Minhan Park and Licheng Yi demonstrate rubric authoring with Wayfinder in MAPLES

Publication scope

Project information and this technical report are available at https://github.com/JamiesonLabUTSW/oasis and https://jamiesonlabutsw.github.io/oasis/. Application and component source code, binaries, installation materials, sample data, and tagged software releases are not part of this publication. Any future software release will identify its exact contents, version, and terms.