Paper
From Garbage-In/Garbage-Out to Semantic Corpus Remediation: Preparing Institutionally Captured Text for LLMs in Political, Legal, Reputational, and Security Domains
An engineering workflow for transforming institutionally captured political-security text into structured epistemic objects — addressing corpus-level semantic contamination before LLMs are deployed in legal, political, reputational, and defence domains.