Secure document intake


Before a document can answer, it must enter correctly.

Secure intake turns files, scans, correspondence and audio into a source-controlled archive without losing authority, access or retention meaning.

01Receivecustody + manifest
02ExtractOCR + structure
03Classifymatter + sensitivity
04Verifyhash + quality
05Indexrole + retention

Quality controls

  • page count and missing-page detection
  • OCR confidence and manual-review queue
  • tables, handwriting, image and audio boundaries
  • duplicate and near-duplicate handling
  • language and document-type classification

Custody controls

  • source manifest and cryptographic hash
  • matter/entity and sensitivity class
  • access inheritance and explicit exceptions
  • retention, legal hold and deletion owner
  • re-index and rollback record

Deletion must be testable

Removing one source means proving its text, chunks, embeddings, summaries, caches and backups follow the approved policy. “Deleted from the interface” is not an erasure record.

See the readiness decisions that precede intake →

Build the evidence chain at the archive entrance.

Scope formats, custody, quality and deletion before selecting the retrieval stack.

Scope secure intake