OASIS—the Open Assessment and Scoring Infrastructure Stack—connects tools for assessment and feedback. Explore the project and consider how a local setup could support your teaching and research.
Running the application locally and running the AI locally have different hardware and data-handling requirements. Start with the kind of workflow you want to explore.
Exploration and small pilots
A laptop with cloud AI
Run the application on your computer and use an approved cloud model for AI processing.
Plan for Docker, working storage, an internet connection, and your chosen provider. Review what data is sent to that provider.
On-device experimentation
A workstation with local AI
Plan to run the application and compatible AI models on your own hardware.
Memory and accelerator needs depend on the model and on whether you work with notes, transcripts, audio, or video.
A shared team environment
An institutional server
Plan a centrally managed deployment for multiple educators or researchers.
Size compute and storage for your workload, and plan user access, backups, model hosting, and ongoing support with your IT team.
These are planning profiles. The forthcoming hardware guide will distinguish tested configurations from suggested setups and include memory, storage, and accelerator recommendations.