Federal courts 1951 to present. Every holding typed, speaker-attributed, jurisdiction-mapped, and source-traced to the paragraph. Built for AI systems that need to reason about legal authority — not retrieve legal documents.
97.1% pipeline accuracy · human-labeled gold set · 15,000+ items
The documents exist. The structured authority layer does not. That is what we built.
Retrieval-augmented generation returns documents. It doesn't tell you if a holding still applies, whether the jurisdiction agrees, whether what was cited as precedent was dicta, or whether the law moved since the model was trained. Legal authority is not a document. It is an assertion, made by a specific court, in a specific jurisdiction, at a specific time, with a specific precedential weight.
KeyCite and Shepard's are the best tools available today. They were built for human lawyers reading PDFs, not for AI systems that need structured, assertion-level data at query time. We built what comes next.
Chevron deference requires courts to defer to agency interpretations of ambiguous statutes. See Chevron U.S.A. v. Natural Resources Defense Council, 467 U.S. 837 (1984).
A legal AI product asked whether Chevron deference still applies. Below is what a live query against the federal case law graph returns — assertion type, treatment history, affected holdings, and jurisdictional scope.
Coverage: 1st–11th circuits, D.C. Circuit, Federal Circuit
Cases drawn from actual court sanctions orders: Mata v. Avianca, S.D.N.Y. 2023 · Fivehouse v. DoD, D.D.C. 2025 · Johnson v. Dunn, M.D. Ala. 2025
Dissent vs. holding. The distinction most AI systems cannot make.
Every AI system that retrieves legal text can retrieve a dissent. Most cannot tell you it's a dissent. Omniarch tags every assertion with its speaker: majority, concurrence, or dissent. Citing a dissent as precedent is not a subtle error. It is the kind that surfaces in front of a judge.
Every assertion typed as a duty, a right, a privilege, or a liability.
An obligation imposed on a party is not the same as a right granted to one. Omniarch applies Hohfeld's legal typology to every extracted assertion. The type of legal relationship determines what the AI can safely conclude from it.
Three separate clocks. When decided. When effective. When applicable.
When a case was decided. When the statute it interprets took effect. When the rule it establishes applies. These three timestamps are often different. Systems that track only one draw incorrect temporal conclusions. Omniarch tracks all three.
All 13 federal circuits. 1951 to present. Every opinion typed, speaker-attributed, jurisdiction-scoped, and treatment-tracked.
Every citation your AI produces is your word in court. Omniarch gives you the structure to verify it before they ask.
Talk to a Legal AI Expert or reach us at hello@omniarch.inc