On May 19, 2026, MET students gathered for “1 Hour: 3 Tools,” a rapid, insight‑packed workshop led by Sooyoung Yang, Melissa Chung, and Sean Jeon. From personal knowledge systems to AI‑supported study help, each presenter showed not just a tool, but a way of thinking about learning more intentionally.
Here are five takeaways our community can carry forward:
Sooyang Yang’s Takeaways on UBC Helpme Bot
1. Moving Beyond a Search Engine
Shifting the perspective from using AI as a basic search engine to adopting it as an interactive, critical collaborator.
2. Designing the Output
You are the architect of the conversation. Provide specific parameters for the format and the perspective you want the AI to take.
3. Engaging as a Thought Partner and Reviewer
Pasting raw thoughts or draft reflections to ask the chatbot for deep questions to expand your thinking, or to identify the weakest points in the logic.
4. Unpacking Pedagogical Intent
Utilizing the chatbot strategically to understand the “why” behind assignments and connect them back to the course module and literature.
5. Prioritizing Academic Integrity
Checking the sources provided by the tool to cross-verify the facts ourselves.
Sean Jeon’s Takeaways about Osidian
1. Just Start
Do not wait until you fully understand Obsidian. The best way to learn is by writing real notes, testing links, and slowly building your own system. Your first notes do not need to be perfect.
2. Do Not Try to Build the Perfect Structure
A perfect folder system or tag system can slow you down. Start simple, then let your structure grow naturally based on how you actually think, study, and work.
3. Use Time-Blocking to Organize and Review Your Notes
Obsidian becomes more powerful when you revisit your notes. Set aside regular time to clean, connect, summarize, and reflect on what you wrote. Notes become useful through review, not just collection.
4. Make It Your Second Brain, Not a Storage Box
Obsidian is not just a place to collect articles, quotes, and copied information. It should help you process your own thoughts. A good note should include your interpretation, questions, examples, and next actions.
5. Connect Obsidian with AI to Expand Your Thinking
AI can help you summarize, question, reorganize, and expand your notes. Use it as a thinking partner, not as a replacement for your own thinking. The goal is to turn scattered notes into deeper insights.
Takeaways from Melissa Chung’s Notion presentation
1. Leverage the power of databases
The power of Notion lies in the flexibility of its databases. The WYSIWYG editor allows for a naturalistic information architecture (IA). These databases act as a backend for forms wiki pages and can also be shared via web for easy access.
2. Use properties and metadata for retrieval
Start with the properties that are pertinent to your workflow. Don’t feel pressured to utilize metadata and information in templates. Be active with management of IA and prune data that clutters your infrastructure. This is how you can best leverage Notion for effective information retrieval.
3. Manage multiple workflows using database views
Calendar view, table view, list view and many others offer a truly expansive way for you to manage data. Integrate your calendar with wiki links, notes and other project management resources that minimizes friction between separate workflows.
4. Connect Notion to Claude Cowork for a recursive and dialogic retrieval
Build on your workflow with agentic tools like Claude Cowork. MCP connections allow for multiple points of input/output within the Notion/Claude pipeline, also minimizing friction in your information retrieval process. Beware of permissions and security when setting up permissions for agentic LLMs.
5. Notion x Obsidian x Claude for a powerful knowledge management system
Leverage your excellent information retrieval system between multiple tools to build a personal knowledge management system. Notion can also support a knowledge management system since its pages can be linked via wiki links, but consider keeping Notion lean since database management can get wily if not optimized for efficient retrieval (or archival).