Skip to main content

What are Disposition Codes?

After the AI completes a review run and produces its findings, a human reviewer assigns a disposition code — the official decision on the document’s status. The AI provides a recommendation (approve, revise, or reject), but the final call is always made by a human.

Available Presets

Raycaster Doc ships with four disposition presets tailored to different industries:

EPC (Engineering, Procurement, Construction)

CodeNameDescriptionStatus
AApprovedNo comments, work may proceedApproved
BApproved with CommentsMinor revisions, work may proceedApproved
CRevise & ResubmitMajor revisions required, work cannot proceedRejected
DRejectedSignificant issues, resubmit requiredRejected

Life Sciences

CodeNameDescriptionStatus
APPROVEDApprovedDocument meets all requirementsApproved
CONDITIONALConditional ApprovalAcceptable with minor correctionsApproved
REVISERevise & ResubmitRequires significant revisionRejected
REJECTEDRejectedDoes not meet requirementsRejected

BLA Submission (FDA Regulatory)

CodeNameDescriptionStatus
BLA_READYBLA ReadyDocument meets FDA requirementsApproved
MINOR_DEFMinor DeficiencyMinor gaps, acceptable with correctionsApproved
MAJOR_DEFMajor DeficiencySignificant gaps vs. regulatory requirementsRejected
RTFRefuse to FileDoes not meet minimum FDA requirementsRejected

Simple (General Purpose)

CodeNameDescriptionStatus
APPROVEDApprovedDocument is acceptableApproved
CHANGESChanges RequestedRevisions needed before approvalRejected
REJECTEDRejectedDocument is not acceptableRejected

Assigning a Disposition

  1. Open the completed review run
  2. Review the AI’s recommendation and action items
  3. Select the appropriate disposition code from the preset dropdown
  4. The decision is recorded with your identity and timestamp

Re-running Reviews

After addressing action items and making changes, you can re-run the review against the updated documents. Follow-up review runs have access to:
  • The previous review’s action items and findings
  • The previous review’s chat conversation (via read_previous_review_chat tool)
  • The previous version of documents (via download_previous_version tool)
This enables iterative review cycles where the AI can assess whether prior findings have been adequately addressed.