Use case · Researchers

Eyrie for researchers.

For PhD candidates buried in literature reviews, academics juggling three concurrent papers, postdocs cleaning up someone else's lit notes, and independent researchers who can't afford a $30k/year RIS license. Eyrie answers "what have I read that contradicts this?" with a sourced citation — not a hallucinated one.

What researchers need that other apps miss.

Verifiable citations on AI answers.

The thing every other AI notes app gets wrong: when you ask "what have I written about X?" — you need to see which note it pulled from. Eyrie ships every answer with clickable chips to the source files. Click a chip, open the source paragraph. No invented citations, no false authority. This is the difference between AI as research tool and AI as plausible-sounding noise.

Local-first for sensitive work.

Pre-publication research, interview transcripts, IRB-protected data — none of it should live on Notion's servers. Eyrie's storage is plain Markdown on your disk, and AI calls go directly from your Mac to your chosen provider. We never see your notes.

Cross-paper reuse.

Most papers contain ideas you've also written about elsewhere — in a previous paper, a grant application, a teaching note. Eyrie's semantic search across your whole vault means "where else have I argued this?" returns the actual previous text in seconds.

BYO AI — no extra subscription.

You probably already pay for Claude Pro or ChatGPT. Eyrie connects to those (or to your Anthropic API key) so there's no second AI subscription to budget for.

The research workflow.

1. Literature notes as Markdown files. One file per source: paper, book, blog post. Frontmatter for author + year + DOI. Headings for sections you take notes on. Tags for topics.
2. "Concept" notes for evolving arguments. A file per concept you're working on. Wikilinks ([[concept-name]]) to literature notes that bear on it. Smart Categories auto-classify these as "research" — the sidebar groups them together.
3. Daily research log. What you read, what surprised you, what you want to follow up on. Skim once a week to find threads worth pulling.
4. AI as second-pass reviewer. When drafting, ask Eyrie: "What in my literature notes contradicts my current claim?" Get sourced answers. Strongest single defense against confirmation bias.
5. Connect Claude or Cursor for heavier lifting. One-click MCP install. Your existing AI subscription can now read all your research notes — useful for big synthesis tasks, less so for daily writing.

What about Zotero and Obsidian?

Keep Zotero. Eyrie isn't a reference manager. Zotero handles PDF storage, citation generation, and library organization. Use both — Zotero for the bibliography, Eyrie for the writing.

Replace Obsidian if: you're tired of plugin sprawl. Obsidian + Citations + Smart Connections + Templater + Dataview is a working setup, but it's also four maintenance burdens. Eyrie ships the equivalent built-in. Full comparison →

The Carnegie test.

If your last three papers each contained one paragraph that you had to retype from scratch because you couldn't find what you'd previously written — Eyrie pays for itself before your next deadline. If your research process is more "read once, never re-read" — you don't need it.