Engineering
We Made Our Database Deliberately Lose All Its Data
Why we run a production database UNLOGGED on purpose: for a derived, reproducible dataset, durability is a cost, not a virtue, and the recovery path is the refresh path.
How we build legal AI
Most legal AI does not show its work. These are the internals of ours: citation verification, retrieval over US law, native Word tracked changes, and the infrastructure underneath.
Recent
Archive
- Research a Hard Question the Way a Lawyer Would: In Pieces, Then Synthesized
- Knowing When You Do Not Know
- Teaching the Reviewer How Your Firm Negotiates
- Draft the Fix, But Earn It
- Call US Law From Your Own Code: The Statutes and Regulations API
- A Spreadsheet Where Every Cell Is a Cited, Verified Extraction
- The Tracked Changes a Document Is Hiding
- Real Word Redlines, Not a Diff View
- The Law Is Not Clean Data
- Extract Once, Read Many: The Shared Evidence Spine Behind a Matter Workspace
- Sending 'Rewrite This Transcript, NOT a Summary' to the Right Engine
- Catching a Fabricated [N] Before It Reaches the Lawyer
- What Our Citation Verifier Actually Checks
- Hybrid Retrieval for Law, and the Citation-to-a-Subsection Problem