SWIRL 5

One Query, Every Authority: SWIRL 5 Federates Public Case Law and Your DMS

A legal question rarely lives in one system. SWIRL 5 runs a single search across US case law, UK case law, and your own document stores at once, then re-ranks everything into one list you can trace back to source.

← Back to Blog

Ask a litigator where the answer to a research question lives and the honest reply is "in several places at once." The governing authority might be a recent High Court judgment. The persuasive authority might sit in a US appellate opinion. The facts that matter to your matter sit in the engagement letter, the draft motion, and the correspondence held in your document management system. Today most teams run that one question three or four times, in three or four tools, and stitch the answers together by hand.

SWIRL 5 collapses that into a single query. Below is a live search for reasonable person, scoped to a litigation workspace, running at the same moment against US case law from CourtListener, UK case law from the National Archives, and the firm's own files in iManage and Box.

SWIRL 5 results for the query 'reasonable person' showing re-ranked results from UK National Archives case law, US CourtListener case law, iManage and Box, with a Sources panel counting hits per source.
One query, four sources, one ranked list. The Sources panel shows how many results each connector contributed; every row carries its origin and a link back to the authority.

Public case law is a first-class source

SWIRL 5 treats public legal corpora the same way it treats your internal repositories. CourtListener gives broad coverage of US federal and state opinions; the National Archives publishes the official record of judgments from the courts of England and Wales. Both are queried in parallel, with their own result counts shown in the Sources panel, so a researcher can see at a glance that ten US and ten UK authorities came back alongside the firm's own work product.

Your DMS stays exactly where it is

The same search reaches into iManage and Box without copying a single document. SWIRL queries each system live, using the searcher's existing permissions, and returns pointers to the originals; it does not ingest, index, or move the files into a vector database. For legal teams that means the matter file keeps its access controls, its audit trail, and its retention policy intact. Nothing is duplicated into a store that your records team never signed off on.

One ranked list, not four inboxes

Federating the sources is only half the job. SWIRL 5 then re-ranks every result, regardless of where it came from, into a single relevance-ordered list. A UK judgment, a US opinion, and an internal draft compete on the merits of the query rather than being penned into separate tabs. The result you see at position one is the most relevant across all four systems, not the best result from whichever tool you happened to open first.

Every row is traceable

Each result names its source and links to the authority behind it: the National Archives citation, the CourtListener opinion, or the document in your DMS. There is no synthesized summary standing between you and the record. When the answer feeds an AI assistant or an agent over SWIRL's MCP server, those same citations travel with it, so a generated answer can always be checked against the opinion or the file it rests on.

Why this matters for legal work

Defensibility in legal research is about showing your work. A federated query that spans public case law and the firm's own files, returns one ranked list, and keeps every result tied to its source is the difference between an answer you can cite and an answer you have to re-verify. SWIRL 5 is built so the citation is never lost on the way to the answer.

Want to see SWIRL 5 run against your own case law sources and document stores?

Join the Preview