A planner, a researcher, a writer, a verifier, and a human reviewer. Bounded scope. Logged actions. Predictable outcomes.
From the docket entry to a partner-ready first draft, with every citation independently checked.
Planner ingests pleadings, prior orders, and the disputed discovery requests.
Issue list and section outline, with proposed authority per argument.
Writer renders in firm voice and chosen style guide. Bluebook formatted.
Independent verifier checks every cite, quote, and claim against the source.
You redline. Litigo files the audit record alongside the draft.
$ litigo run depo.summary --transcript ./smith-depo.pdf --witness "Smith, J." \
--matter HARTWELL-v-CARDIN
[ 00:00:02 ] planner → 412 pages ingested · 1,184 Q&A turns parsed
[ 00:00:14 ] extractor → 38 admissions · 12 contradictions · 7 follow-ups
[ 00:00:21 ] writer → summary.md (3,402 words) · timeline.md · prep.md
[ 00:00:24 ] verifier → 47/47 citations match transcript line numbers
[ 00:00:25 ] ✓ ready for review · runtime 25.1s · cost 412 credits
Litigo's outcome model is trained on millions of state and federal dockets — case posture, judge tendencies, motion timing, prior holdings. It returns a calibrated probability band, not a single number, and shows you the comparable matters it leaned on.
Modeled on 1,408 comparable matters before Judge Whitfield, S.D.N.Y., 2014–present.
Compel · dismiss · summary judgment · in limine · sanctions.
Requests, responses, privilege review, hot-doc surfacing.
Summary, contradiction map, prep pack, exhibit list.
Issue spotting → controlling authority → memo, with cites verified.
Posture analysis, settlement bands, motion sequencing.
Plain-English matter status, auto-drafted, partner-approved.