Meet DARIA, the cutest little AI that's about to make grading disappear!

I will demo the application and code for DARIA (Data-driven Assessment for Real-world Inquiry and Application), a prototype tool focused on making high-quality, project-based learning scalable for technical instructors.

The demo will focus on the technical implementation of the multi-modal data pipeline, which handles complex submissions (e.g., a .md or .pdf containing code, outputs, and visualizations) and prepares them for assessment. I will show the strategies for fine-tuning a specialized multi-modal model to handle this combined input, and discuss the methods used for securing the model while providing structured, rubric-aligned output. This is a deep dive into the code behind building an effective, complex assessment engine.