Obsidian is a local-first note-taking app built on plain markdown files with bidirectional linking. When combined with AI tools, it becomes a personal knowledge system where your past reading and thinking actively help you with current work.
Why Obsidian for Research
Most researchers write reading notes that disappear into folders never read again. Obsidian’s bidirectional links and graph view reveal connections between notes — a note on “methodology X” automatically links to every paper you’ve read that used it. After 6–12 months of consistent note-taking, the graph becomes a queryable map of your knowledge that accelerates new literature work.
The Core Workflow
For each paper you read: create a note named by author-year-keyword (e.g., “Smith-2023-transfer-learning”). Include a 3-sentence summary, key claims with page numbers, methodology notes, and connections to other notes using [[double brackets]]. Write your own reaction: “This contradicts Schmidt-2022 because…” and “Could apply to my chapter 3 because…” Over time, the reaction notes contain your developing scholarly thinking.
AI Integration Points
Several Obsidian plugins connect to AI: Smart Connections (finds semantically similar notes when you open a new one), Text Generator (generates text within Obsidian using OpenAI/Claude APIs), and Obsidian Copilot (chat interface connected to your vault). These let you ask “What have I read about X?” and get answers from your own notes rather than from AI training data.
The Zettelkasten Method
The most productive long-term structure for research notes is Zettelkasten: each note contains one idea, ideas connect to related ideas, and the system grows by adding new idea notes and new connections. Niklas Luhmann developed this method and wrote 70+ books using it. For PhD researchers, starting a Zettelkasten in year 1 and maintaining it consistently produces a compounding intellectual asset by years 3–4.
Syncing and Backup
Obsidian vaults are plain markdown files — sync with iCloud, Dropbox, or Obsidian Sync. Back up to a private GitHub repository for version control. This setup means your knowledge base survives any device failure and has a full edit history.
