Skip to main content

Overview

When you upload a contract to Smart Drive, it goes through a multi-stage AI processing pipeline that extracts key deal terms, identifies parties, and structures the data for use across Royaltyport. Processing Pipeline

Processing Stages

1. Preprocessing

The initial stage where documents are prepared for AI analysis.

Queued Documents

Documents waiting to be processed appear with a Queued status. The queue processes documents in order of upload, with some prioritization for smaller files.
InfoDescription
Queue PositionYour document’s place in the processing queue
Estimated TimeApproximate time until processing begins

Document Preparation

During preprocessing, the system:
  • Converts file formats to analyzable text
  • Runs OCR on scanned PDFs and images
  • Identifies document structure and sections
  • Prepares the document for AI extraction

2. Processing

The main AI extraction phase where deal terms are identified and structured.

Initial Details

The AI first extracts basic document information:
FieldDescription
Contract TypeRecording, Publishing, Distribution, etc.
LanguagePrimary language of the document
Page CountNumber of pages in the document

Contract Terms

Detailed extraction of deal terms:
  • Financial terms (advances, royalty rates, accounting)
  • Dates (execution, term, expiration)
  • Territory (geographic scope)
  • Rights (what rights are granted)

Contract Text

The full text is indexed for:
  • Search functionality
  • AI chat queries
  • Reference linking

Contract Indexing

The document is indexed for fast retrieval:
  • Full-text search
  • Semantic search capabilities
  • Cross-reference with other contracts

Processing Statuses

StatusDescriptionAction
QueuedWaiting in the processing queueWait for processing to begin
ProcessingAI is actively analyzing the documentWait for completion
ReadyProcessing complete, data availableReview extracted data
FailedProcessing encountered an errorCheck document, retry, or contact support
PausedProcessing paused, requires inputProvide required information

Failed Documents

When processing fails, the contract shows a Failed status.

Common Failure Reasons

ReasonDescriptionSolution
UnreadableDocument text cannot be extractedUpload a clearer version
CorruptedFile is damagedRe-upload the original file
UnsupportedFormat not supportedConvert to PDF or DOCX
Password ProtectedDocument is encryptedRemove password and re-upload

Retrying Failed Documents

1

Open the Document

Click on the failed document in the Drive.
2

View Error Details

Check the error message for specific issues.
3

Fix the Issue

Address the problem (e.g., upload a clearer scan).
4

Retry Processing

Click Retry to reprocess the document.

Paused Documents

Processing may pause when user input is required.

Why Processing Pauses

ReasonDescription
Split ConfirmationDocument may contain multiple contracts
Entity VerificationNeed to confirm party identification
Ambiguous DataAI needs clarification on extracted values

Changing/Adding Splits

When the AI detects multiple contracts in one document:
1

Review Suggested Splits

See where the AI suggests splitting the document.
2

Adjust Split Points

Modify the page ranges if needed.
3

Confirm Splits

Accept the splits to create separate contract records.
4

Or Keep as One

Reject splits to process as a single document.

Resuming Paused Processing

1

Find Paused Document

Filter by “Paused” status in the Drive.
2

Provide Required Input

Enter the information requested.
3

Resume Processing

Click Continue to resume extraction.

Processing Time

Expected processing times by document type:
Document TypePagesTypical Time
Simple contract1-530 seconds - 1 minute
Standard contract5-201-3 minutes
Complex contract20-503-5 minutes
Large document50+5-10 minutes
Processing times may vary based on current system load and document complexity.

Real-Time Updates

The Drive shows real-time status updates:
  • Progress indicator shows extraction progress
  • Status changes appear automatically
  • No refresh needed - updates stream in real-time

Batch Processing

When uploading multiple documents:
  • Documents are queued in upload order
  • Processing happens in parallel where possible
  • Each document’s status updates independently

Next Steps