Method
Search terms retrieve records. They do not prove meaning.
Retrieve
State credit-card and checkbook records were screened across 91 topics.
Merchant strings, agency names, destinations, and policy-related terms were used to locate candidate rows—not to predetermine conclusions.
Separate
Mixed files, false matches, ordinary spending, and rows needing records were kept distinct.
The exhibit structure preserves the difference between a facial contradiction and a transaction that may be explainable once receipts or approvals are produced.
Publish
The strongest public-facing findings were consolidated into a release package.
The package includes the public narrative, QC memorandum, exhibit index, top findings, school-sector summaries, and underlying screening layers.