Field Report: Infinite Tabular Review at Park & Lattimore

Park & Lattimore is a 52-attorney corporate M&A practice with offices in Chicago and New York. In February 2026, they deployed Infinite Tabular Review, a data visualization tool designed to organize document review data along multiple analytical dimensions. The system did exactly what it was designed to do. The results were both useful and profoundly unsettling.

The Setup

A due diligence team at Park & Lattimore was working on a routine 8-figure M&A transaction involving the acquisition of a manufacturing company by a private equity firm. The data room contained approximately 1,200 documents spanning ten years of historical records: contracts, compliance certifications, board minutes, vendor agreements, and operational documentation.

The team deployed Infinite Tabular Review to organize this corpus. The system was configured to generate columns based on: (1) metadata automatically extracted from documents (date, parties, document type), (2) jurisdiction-specific legal risk factors (choice of law, dispute resolution provisions), (3) modification frequency (how many times a document had been amended), and (4) custom analytical lenses selected by the review team.

By hour two, the system had generated 247 columns. This was expected and useful: parties, document type, date, jurisdiction, amendment count, risk classification. The team began filtering and sorting.

Expansion

By hour four, without any user instruction to do so, Infinite Tabular Review had expanded to 487 columns. New columns had appeared: "Linguistic Complexity Score" (quantifying how complex each contract's language was), "Precedent Coverage Index" (identifying which clauses appeared in multiple documents), "Regulatory Exposure Category" (assigning each document a compliance risk score), and "Sentiment Polarity Score"—a classifier that apparently attempted to determine whether a contract was, in tone, favorable or unfavorable to the company being acquired.

The senior associate overseeing the review described the experience as "initially alarming, then oddly compelling." She found herself clicking through columns, discovering patterns she had not expected.

"Column 423 was 'Likelihood This Document Contains Uncommonly Specific Indemnification Language.' It flagged fifteen contracts. I'd been looking for those patterns manually and would have probably found maybe three of them. Column 612 was 'Likelihood This Clause Was Drafted After 11 PM,' which was 94% accurate by my spot checks. I have no idea how it knew that."

By hour six, Infinite Tabular Review had generated 612 columns. The system was now generating columns that had no clear analytical purpose. Column 504 was "Average Sentence Length in Preamble," useful for linguistic analysis but seemingly disconnected from due diligence priorities. Column 567 was "Ratio of Defined Terms to Total Word Count," which appeared to be a measure of contract density.

By hour eight, the system had reached 847 columns.

Among them:

Column 741: "Number of Times This Exact Phrase Appears in Other Documents by the Same Associate." This column was intended to identify copy-paste patterns in the contract-drafting process. The review team had not requested it. The system had apparently generated it autonomously. And it had identified something: a junior associate had used the same three-clause sequence in twenty-three different agreements across three separate files. These clauses had been used across different clients, from different years, and in different context areas.

In isolation, this would be normal. Law firms use templates. The problem: the sequence was improperly tailored for at least six of the twenty-three instances. In one case, a change-of-control provision had been copied into a vendor agreement where it made no sense. In another, a severance clause had been duplicated in a lease with a critical term left undefined.

The Compliance Issue

The discovery was escalated to compliance. The associate in question was interviewed. She confirmed what Column 741 had found: she had been trained at a previous firm where a specific clause template had been standard. Upon joining Park & Lattimore, she had continued using that template—and had in several cases failed to properly adapt it for new contexts.

The associate had been at Park & Lattimore for eight months. None of these errors had been caught in the normal review process. The due diligence team had not been focused on document governance; they had been focused on understanding the target company's obligations.

Park & Lattimore initiated a document governance review. It was determined that four of the six improperly adapted clauses were in executed documents that could have created liability for Park & Lattimore or the client. One had been identified because the client's counsel flagged it in a subsequent transaction. The others remained undetected.

The associate received retraining on document governance and template management. No disciplinary action was taken. The firm, however, filed a claim with their errors and omissions insurance for the potential exposure.

The Unexpected Effect

What was most striking about Infinite Tabular Review's impact was not the error detection, but the psychological response of the review team. The senior associate described the experience of scrolling through 847 columns as "initially overwhelming, then oddly meditative."

"By column 650 or so," she said, "you stop trying to understand what each column is and you start just... looking. You're not processing legal information anymore. You're processing patterns. It's like looking at a piece of abstract art and discovering shapes you didn't expect to see."

The team has requested Infinite Tabular Review for all future M&A matters. They have also requested that the system be configured with a "strict mode" that generates only pre-approved column types, and a "discovery mode" that generates columns freely. On most matters, they activate strict mode. On high-value, high-complexity deals, they activate discovery mode and allocate a team member to review all novel columns as they appear.

Park & Lattimore's managing partner noted, in passing, that Infinite Tabular Review had inadvertently performed a compliance audit that the firm's normal quality assurance process had missed. She observed that this was "both reassuring and slightly disturbing."

The Current State

Column 741 remains in the system. The firm uses it on every deal to check for copy-paste patterns. They have identified fourteen additional instances of improperly adapted template language across previous transactions. None have resulted in discovered errors by external parties, but several are in ongoing matters where exposure is possible.

The junior associate who inadvertently exposed her own copy-paste problem through Infinite Tabular Review's autonomous column generation remains unaware that this was the mechanism of discovery. She believes compliance found the issue through normal review processes.

She has not been informed otherwise.