Ask the file. Get the page.
Ask questions of a case’s documents and get answers with a citation to the exact page behind every statement. Click a citation and the file opens at that page.
Drafts that stream into Word in your firm’s voice, grounded in the case record, with a numbered badge behind every sentence taken from a document. Click the badge and the file opens at that page. The same citations run through Case File Chat, Discovery Sets, Extractions, Medical Chronologies and Discovery Responses.
In the task pane, choose the case and the firm’s document type. Its house style and standing instructions load automatically, and the choice is saved in the .docx so a reopened document knows what it is.
Attach files from the case folder, or point the AI at an indexed discovery set so it searches hundreds of documents for what the draft needs.
The draft lands at the cursor, or rewrites what you selected, in the firm’s voice. Formatting is applied as it lands. Every sourced sentence gets a numbered badge that opens the file at that page.
Unsupported sentences are highlighted in place with Rewrite to the sources. Every case cited is looked up on CourtListener. Hide the badges so they never print, or remove them all in one click.
Ten capabilities in the task pane, all of them working on the document that is already open. Nothing leaves Word, and nothing gets pasted.
Type what you need and the draft streams straight into the document at the cursor, or rewrites the text you selected, in the firm’s voice and grounded in the case record and the files you attach. Formatting is applied as it lands, so the text is ready to keep, not paste.
Every passage drawn from a document gets a small numbered badge at the end of the sentence. Click a badge and the source file opens at that page. The citation list travels inside the document, badges can be hidden so they never print, and one click removes them all.
Pick the firm’s document type in the task pane and its house style and standing instructions load automatically. The choice is saved in the .docx, so a reopened document knows what it is.
Pick the case and attach files from the case folder, or point the AI at an indexed discovery set so it searches hundreds of documents for what the draft needs.
For a sectioned document type, the AI proposes an outline for this case, you check the sections you want, and it writes them in order under proper headings. Any section can be redone on its own.
Checks the selection or the whole draft against the attached files and the case record. Unsupported sentences are highlighted in place, with Rewrite to the sources or Add badge for a sentence the sources do support.
Looks up every case cited in the draft on CourtListener, a public database of court opinions. Cites that no case is reported at, or that belong to a different case, are highlighted red; ones that cannot be looked up are amber. Only the citation is sent, never the document. Confirms a case exists; not legal research.
A chronology built from the case’s records inserts into the document as text with a citation badge on every entry, or attaches to the AI as context for the draft.
A reviewed set of discovery responses assembles in Word in the firm’s format: preliminary statement, objections, each request with its response and source badges.
The last instructions you gave the writer are one click away, with the mode and document type they ran under.
Everything the AI knows, it learned from the case’s own documents, and every answer says which page it came from.
Ask questions of a case’s documents and get answers with a citation to the exact page behind every statement. Click a citation and the file opens at that page.
Index a case folder once and the AI searches hundreds of documents for what a question or a draft needs, instead of reading one file at a time.
A dated timeline of treatment across a set of medical records, merged into one chronology with a citation behind every entry. Gaps and missing records are flagged.
Ask one question of every document in a set and get the answers as a table, one row per document, each with cited quotes. Export to Excel.
Each document type carries the firm’s house style, learned from its own samples, plus standing instructions. Nine starter types are one click away.
Reads a served demand, splits it into numbered requests, answers each from the case file with the firm’s own objections, and assembles the response in Word after you review it.
Every sentence is checked against the files and the case record. Every case cited is looked up on CourtListener. Problems are highlighted in place.
Ask questions of a case’s documents and get answers with a citation to the exact page behind every statement. Click a citation and the file opens at that page.
Index a whole case folder so the AI can search hundreds of documents for what a question or a draft needs, instead of reading one file at a time.
Ask one question of every document in a set and get the answers as a table, one row per document, each with cited quotes. Export to Excel.
Every firm-defined document type carries the firm’s house style, learned from its own samples, plus standing instructions, so a deposition summary or demand letter comes out the way the firm writes it every time.
Nine ready-to-use types an office can add in one click: deposition summary, position summary and carrier report, demand letter, client update letter, medical chronology, discovery responses, complaint, motion, and bill of particulars. Each is editable and rebuilds from the firm’s own samples.
A document type can define its sections. The AI plans an outline for the specific case, drafts each section in sequence with the earlier ones in context, and streams the whole thing into Word with proper headings.
Builds a dated timeline of treatment across a set of medical records, merged into one chronology with a citation to the page behind every entry, with gaps in treatment and missing records flagged.
Reads a served demand, splits it into its numbered requests, answers each one from the case file with the firm’s own objections, and assembles the response in Word. Every answer is reviewed and edited in a table before anything is inserted.
The firm’s standing discovery objections, in the firm’s exact wording, organized by state. A standard set for any of the 50 states, the District of Columbia, or the federal rules can be added in one click and edited to match the firm’s practice.
Draft a reply to a case email from inside the case, in the firm’s voice, with the relevant files as context, without leaving the compose window.
A per-office log of every AI run: who ran it, when, which feature, which case, which documents it read, and what it cost, with usage totals by member and by feature and a spreadsheet export.
A plain-language statement of how the AI handles a firm’s documents: no training on firm data, zero retention at the model provider, prepared copies kept only in the firm’s own storage, and an audit trail for every run.
A plain-language statement of how the AI handles your firm’s documents, backed by executed agreements with every model provider, available to administrators inside the platform.
Read the full statementBring a real file. On the demo we’ll index it, chat with it, build the chronology and stream a draft into Word, with the citations, before the call ends.
No credit card required · Free onboarding · Cancel anytime