Organising your Research Notes Efficiently on Protolyst

A Protolyst workspace can turn familiar Word-style research templates into a structured database that automatically extracts citations, converts highlighted snippets into labeled “atoms,” and then displays those atoms across filterable tables. The payoff is less time hunting through PDFs and more time recombining notes—by keyword, by question category, or by combinations of both—without rebuilding everything in a separate system.

Setup starts in a document-style table. After adding a table, the table type is switched to “Vertical View,” which effectively transposes the layout so that what would normally be columns become row titles. A title is added to anchor the paper being uploaded, and headings are recreated to match the existing Word template. From there, key Protolyst properties are added to the table: a “citations” property that auto-populates reference details when a PDF is uploaded (with selectable citation styles), and an “atoms” property that uses filters tied to the template’s questions. Those filters matter because any atom dragged into the atoms field inherits the corresponding label (for example, background information) automatically.

The template also becomes a tagging and media capture system. Global tags aren’t handled as simple labels; instead, they’re created as pages so they can be applied across the workspace. For each question category, new pages are created to match the template’s structure. An “image” property is added so figures can be captured and stored alongside notes, and a “multi select” property is used to organize papers by keywords (topics). With this in place, each row in the table functions like a paper record: upload a PDF and Protolyst pulls in the title and citation information automatically.

As reading begins, the workflow shifts from manual copying to drag-and-drop extraction. When a user highlights text in a paper, a capture atom option appears; instead of copying text into a separate document, the highlighted content can be dragged into the relevant atoms property. Because the atoms field is filtered, the atom is automatically categorized and labeled. If figures are needed, screenshots can be pasted into the figures image property. Each captured atom can also include annotations—rewriting the sentence in one’s own words or adding new ideas—stored alongside the atom text.

Once multiple papers are added, the workspace becomes navigable in two complementary ways. In document-style view, the table shows each paper’s properties and the categorized atoms, letting researchers scroll through notes similarly to a Word document. In table view, specific properties can be displayed or hidden, and filters can be saved as reusable views. Keyword filters (from the multi select property) can show only papers tagged with a term like “hydrogels” or “3D printing,” and the logic can be adjusted to include or exclude papers based on tag presence.

Finally, the system supports writing without reference hunting. Atoms collected from relevant papers can be dragged into a writing page; Protolyst carries over atom text and any annotations, and it automatically creates citations by pulling reference details from the corresponding paper’s citation property. The result is a research workflow where notes are continuously structured, searchable, and ready to recombine into new drafts.