Thu, Oct 23 · 7:00 PM HST
This is a free, two‑hour, hands‑on session focused on building practical chatbot prototypes that connect to Canvas. We’ll start with a short overview of Model Configuration Protocols (MCPs) and how a simple MCP can talk to the Canvas/Instructure API to answer real questions about assignments, due dates, and course activity.
Then we’ll walk through a live demo of a working MCP and invite participants to follow along in Replit, with plenty of time for guided tinkering and Q&A.
This meetup is for K‑12 and higher‑ed technologists, classroom teachers, school leaders, and curious parents who want to see what’s possible—and how quickly a small team, or even a single teacher, can stand up useful tools. The goal is to empower Hawaii’s education communities to prototype and deploy lightweight assistants that help families and students get on the same page about classroom work in Canvas. We’ll also discuss how the same approach could extend to other platforms and SIS/LMS ecosystems.
A quick note on privacy: We will model best practices for secure, FERPA-aware development. The approach we'll focus on is building "personal assistants," where a parent or student uses their own personal API key to access only their own information. This method empowers users directly while respecting student privacy. For the workshop, all activities will use a safe, sandboxed demo account with fictional data, and we'll publish starter code after the meetup so you can keep building.
Laptops are welcome. If you have Canvas access, try to generate a personal access token ahead of time and know your Canvas domain (for example, yourschool.instructure.com) so you can connect during the session. If you don’t have a token, you can still follow along; we’ll provide a demo path in Replit.
Basic familiarity with code helps but is not required, and we’ll keep the pace inclusive and collaborative.
We’ll meet at the Manoa Innovation Center on Oʻahu. If participants register from neighbor islands, we’ll add a Zoom option and share the link with registrants before the event. Feel free to comment on the event page or email questions in advance, and we’ll adapt the plan based on interest and needs. After the meetup, we’ll publish starter code and links so you can keep building.
By the end, you’ll have seen a working Canvas‑connected chatbot via MCP, tried a prototype yourself, and left with clear next steps to continue on your own or with your school.
