Derived from a 2.9-million-row forensic database built from 4,251 court filing PDFs spanning 2,903 cases across Minnesota's Fourth Judicial District
In January 2023, Matthew Guertin was arrested on four firearms charges. Within four days, a competency evaluation was ordered. Three successive court-appointed evaluators classified his patent theft claims as delusional — without verifying the publicly searchable, already-granted US patent (11,577,177 B2) that sits at the top of Netflix's own patent references.
In response, Guertin built a forensic database. He downloaded 4,251 court filing PDFs from the Minnesota Court Records Online (MCRO) system, extracted their metadata, parsed their internal objects, and loaded everything into a 53-table PostgreSQL database. What emerged was a pattern no one was supposed to see.
Every report is derived from SQL queries against the forensic database — not opinion, not interpretation, but measured outputs from authenticated court records. The database itself is cryptographically verified: 98.9% of court PDFs carry valid digital signatures, cross-referenced against Bitcoin blockchain timestamps via a custom hash-chain evidence capture system.
The reports span 13 analytical categories — from hearing transcript claim verification to font forensics, from mental health statistical patterns to patent duplication analysis. Together they constitute the most comprehensive forensic audit of a single county's competency evaluation pipeline ever assembled by a criminal defendant.