Every figure, to a page and line.
Built for plaintiff toxic-tort teams. Each firm works in its own private workspace; documents come in from a local folder or any shared link, read-only.
TortPlus reads the text layer of deposition transcripts, finds the exposure testimony, and turns it into a working evidence desk: timeline, defendant matrix, evidence graph, alerts and reports, every figure traceable to a page and line.
Drop transcript PDFs. Proposed exposure events, cites and anomalies land in a review queue before anything enters the case.
Each bar is one exposure event as the deponent dated it — product, site, years — with the testimony that supports it.
One row per defendant, assembled from linked exposure events. Enter an exposure once; it flows to every defendant it supports.
Deponent → exposure events → defendants. Isolate any node's evidence path; see at a glance what has no cite behind it.
Rules run over the graph on every edit: self-conflicting testimony, products identified but unlinked, defendants with nothing behind them.
Plain-language questions over the case and your linked folders and repositories. Answers cite [Doc n, p. x] and transcript pages with live links; a ledger records every read.
Name the case and deponent, drop the transcript PDFs. TortPlus reads the text layer and drafts cited exposure events.
Accept, edit or reject each proposed event and cite. Nothing enters the case until someone on the team says so.
Tick the defendants each exposure supports. The matrix, timeline and graph recompute as you go; alerts flag what is thin.
Question the case in plain language, then generate the exposure report — Markdown or HTML, every figure traced to its cite.
TortPlus is built around attorney work product. The case lives in each user's browser; documents are read where they already live, read-only; the model sees a question and the excerpts needed to answer it, nothing more.
Shared cases: every case is saved to the firm workspace as you work, and any member opens it from any browser, with versioning and a receipt for every save. Scanned transcripts are read with OCR in the browser.