SquareLog

A discovery binder with 40,000 pages doesn't have a reading comprehension problem—it has a retrieval problem. The answer is there. Your brain isn't.

In litigation, the documents that matter are buried in noise. Depositions, expert reports, exhibit indexes, privilege logs, chain-of-custody records. A junior associate can read for six hours and miss the exhibit that contradicts opposing counsel's claim. SquareLog keeps you from being that junior associate.

What it is

SquareLog is a legal-evidence RAG system purpose-built for discovery workflows. It handles the kind of corpora litigators actually deal with: heterogeneous formats (transcripts, PDFs, scanned documents, redaction schedules), metadata that matters (privilege flags, attorney work-product designations, dates, party names), and the audit requirements that come with federal rules of evidence.

Unlike generic RAG systems that treat all documents the same, SquareLog understands legal privilege. It respects attorney-client privilege, work-product doctrine, and privilege-adjacent designations. Every retrieval leaves an audit trail—who asked what, what the system returned, when. That trail survives a privilege-log challenge.

The system grounds citations back to the source documents. No hallucinations. No invented exhibits. If it says "see Deposition Tr., p. 47," it means page 47 exists and says what you asked about.

What it does well

Privilege-aware retrieval. SquareLog doesn't treat privilege as a checkbox. It understands that documents can be sensitive in different ways—attorney-client communications, work product, trade secrets, settlements with confidentiality agreements. It flags them in results, respects them in filtering, and logs every touch for your privilege log.

Context-rich search. The system understands that a date range, a party name, and a phrase aren't separate queries—they're one question asked with more precision. Filter by date, parties, document type, or privilege status, then ask your semantic question inside those bounds. No false positives from unrelated cases or irrelevant document types.

Deposition intelligence. Deposition transcripts are cheap to store and expensive to search manually. SquareLog indexes testimony by speaker, date, subject, and claim. Ask it "what did the defendant say about the safety protocol," and it returns the relevant portions in context, not just relevance scores.

Exhibit handling. Exhibits are often the point. SquareLog tracks exhibit numbers, dates, and relationships. When you ask about an exhibit, it returns the exhibit, its metadata, and the testimony or document that references it—in one retrieval, not three separate searches.

Audit-ready logging. Every query, every result, every filtered-out document is logged with timestamp, user, and explanation. Export your retrieval log. It survives discovery. No "I don't remember what I searched for."

Who it's for

Small to mid-size law firms handling document-heavy cases. Firms without an IT department but with cases where discovery is the bottleneck. In-house counsel managing large document collections. Litigation support teams that spend their afternoons in PACER downloads and Excel pivot tables trying to organize what they've already read three times.

Specifically: litigators who bill hourly and would rather bill brain than clock. Associates preparing expert reports. Counsel reviewing privilege logs before production. Partners who need to move faster without hiring more junior associates.

What it looks like in practice

A partner is defending a breach-of-contract claim. Opposing counsel claims your client made explicit promises about support timelines. You have 23,000 pages of email, chat logs, and statements of work. You could assign an associate to read. Instead, you ask SquareLog: "What did we promise about support response times, and when?" The system returns the relevant passages from three specific emails and two statements of work, with dates and page numbers. You have your answer in three minutes.

Later, opposing counsel produces a deposition transcript claiming the defendant testified to an even broader promise. You ask SquareLog: "Find the parts where the defendant talks about support timelines." It returns six passages, in context, with speaker name and page number. You have the contradiction. You cite it.

What it's not

SquareLog is not a case-management system. It doesn't track deadlines, manage witness lists, or generate pleadings. It's not OCR—you provide clean documents or handle OCR yourself. It's not a replacement for thorough review; it's a tool for finding what you've already decided to look at.

It's not designed for the kind of document management where you're trying to organize thousands of overlapping cases. It works best with defined collections—a single matter, a deposition, a privilege log.

The mobile companion

square-log-mobile is the local-first mobile capture surface for the SquareLog ecosystem — built so a parent, paralegal, or litigant can log a time-stamped, objective SitRep entry in seconds, from a phone, with or without a connection. Captures sync back to the main SquareLog instance when connectivity returns. .NET MAUI Blazor Hybrid, SQLite + SQLCipher local-encrypted, with on-device OCR.

Status

Shipping. Talk to us for a walkthrough.