Appellate & family law

Case work, without limits.

Source-grounded AI for lawyers handling the record — from message exports and bank statements to briefs and filings.

Every answer cites your evidence
File through your e-filing provider
No training on your matter data

Structured extraction pipeline

PDFs in. Facts in Postgres.

Upload once. Python classifies, routes, and extracts structure. Everything downstream — chat, timeline, drafts — reads persisted rows, not raw PDFs.

Pipeline

Upload

File hits the queue.

A PDF lands in your matter. Storage writes the bytes; BullMQ hands the job to the ingest worker.

GCSBullMQingest worker
exhibit_packet.pdf
01

Upload

02

Classify pages

03

Route segments

04

Extract structure

05

Persist

06

Work the matter

Product

Every surface. One case file.

Chat

Ask the whole matter. Get answers with page-level citations — contradictions, hearing prep, or draft strategy grounded in your evidence.

Read more

Timeline

Texts, transfers, filings, and hearings on one chronology tied to source documents — not a paralegal spreadsheet.

Read more

Documents

iMessage exports, Zelle and bank records, email threads, court forms, and scanned filings — organized by matter.

Read more

Case

Theories, contradictions, drafts, record designation, and filing — from the first upload to the opening brief.

Read more

Security

Built for the confidentiality demands of litigation.

Authenticated workspaces. No model training on your matter data. Privileged material stays in the case file — not a public chat box.

Read privacy policy

No training on case content

Your uploads are never used to train models.

Workspace-scoped access

Authenticated sessions with matter-level isolation.

Source-grounded outputs

Answers and drafts cite evidence in your file.

E-filing ready

Submit through your configured filing provider.

Discover the future of case work through a live demo.

See chat with citations, timeline analysis, and drafting — from upload to filing.