I’m developing and testing a custom AI Study Companion for my course at Parsons as part of The New School’s AI Course Study Companion Pilot, working with The Innovation Center, The Faculty Center, and the XReality Center.
The goal for my course is very specific: help students get more out of the college art and design portfolio process as they prepare to apply to the schools of their choice.
The pilot team developed the overall framework and AI prompt structure. I’m adapting that framework to my course—building the knowledge base, shaping how the AI responds, testing different kinds of student questions, and setting clear boundaries around what the AI should and shouldn’t do.
I built the source collection around the actual portfolio process. It combines my course materials with selected resources on developing and selecting work, sketchbooks and process, photographing and documenting work, digital presentation and sequencing, and personal statements. I also added current portfolio requirements from Parsons and other art and design schools so students can research and compare what individual schools actually require.
Getting my own course material into the AI also required some problem solving. Since my weekly Canvas pages and assignments are behind the Canvas login, I converted them to PDFs and added them directly to the source collection along with the syllabus.
I’m now testing the Study Companion with questions students actually ask: How many pieces does a particular school require? Do they want to see sketchbook and process work? What should I include? How should I photograph and present my work? Testing has also helped me refine how the AI uses sources, cites specific requirements, and avoids treating one school’s requirements as a general rule.
The next step is using the Study Companion with students throughout the course. I’ll be looking at where it helps, where it doesn’t, and what needs to be adjusted. What we learn from that work will contribute to the larger New School pilot at the end of the semester.