Every number here is live.
Every digitized opinion, every docket, and every section of all 137 NY Consolidated Laws — pulled from CourtListener’s bulk dumps and the NY Senate OpenLegislation API, embedded with mxbai-embed-large, and stored in AWS Aurora PostgreSQL alongside the relational citation graph. The counts below are fetched the moment you load this page; while the server tallies them, watch the digits spin.
Case decisions by court
What lives where
The source-of-truth corpus in AWS Aurora PostgreSQL. Every row was streamed in from CourtListener bulk dumps or the NY Senate API and indexed for pgvector semantic search.
The citation graph as relational tables in Aurora. Every NY-to-NY opinion citation is an edge, traversed with recursive CTEs. Pure vector search misses controlling precedent; the graph finds it.
Postgres full-text search for the moment when you already know the citation — the same Aurora engine, across all 137 NY Consolidated Laws.
Live infrastructure
Three ways to use it
Plain-English question, citation-anchored answer drawn from the NY corpus in Aurora.
Semantic case search. Type the legal issue, get the closest NY decisions ranked by citation influence.
Paste a brief and Lex.NY verifies every citation against the real NY corpus in Aurora.
Source: connecting…