Trace for Researchers
Index papers, PDFs, and citations automatically. Ask natural language questions across your entire reading history. Never lose track of a source.
A literature review isn't one session, it's forty tabs open over six weeks, three PDFs re-downloaded because you couldn't remember where you'd already read them, and by the time you're writing the discussion section you've forgotten which paper made which claim. Say you're drafting the related-work section for a paper on retrieval-augmented generation: instead of scrolling back through browser history, you ask "which papers I've read compare RAG to fine-tuning" and get the exact PDFs back, with the passage that made the comparison highlighted. Trace indexes every PDF, abstract, and page you open, full text and an AI summary, and builds a knowledge graph of how the papers relate to each other, so the citation you need is a query away, not a re-read away.
- Full-text search across every PDF, paper, and page you've ever opened, no manual tagging, e.g. "the 2023 paper that benchmarked against GPT-4"
- Knowledge graph surfaces how sources connect: which papers cite the same benchmark, share an author, or make competing claims
- Cite Page skill generates a formatted citation (APA, MLA, Chicago) from any open source in one click
- AI sidebar answers questions across your whole reading history: "summarize what I've read about attention mechanisms this month"
- Works entirely offline with a local model, so unpublished drafts and embargoed papers never leave your device