Document Processing Implementation Checklist

Phase 1: Setup & Upload

• 1. Identify the User

  • Action: Determine if you are serving an existing user or a new applicant.
  • API Call: GET /v1/users/ (Find) or POST /v1/users/ (Create).
  • Success Criteria: You have a valid user_id (e.g., user_12345).
  • Reference: List Users | Create User

• 2. Create the Collection

  • Action: Initialize the document container.
  • API Call: POST /v1/collections/
  • Payload: Must include user_id. Can optionally include all files or some (Base64).
  • Success Criteria: You receive a collection_id.
  • Reference: Create Collection

• 3. Upload Additional Files (Optional)

  • Action: If needed, upload more files to the existing collection.
  • API Call: POST /v1/collections/{collection_id}/files/
  • Reference: Upload Files

Phase 2: Processing & Finalization

• 4. Monitor Document Categorization

  • Action: Poll the collection to check AI processing status.
  • API Call: GET /v1/collections/{collection_id}/
  • Success Criteria: Look for the documents array where status is successful.
  • Reference: Retrieve Collection

• 5. Finalize the Collection

  • Action: Confirm valid documents to generate the Truv Link.
  • API Call: POST /v1/collections/{collection_id}/finalize/
  • Success Criteria: Response includes a link_id.
  • Reference: Finalize Collection

Phase 3: Link Status & Data Retrieval

• 6. Ensure Link Readiness

  • Action: Wait for the Link processing to complete before fetching data.
  • Method A (Webhook): Listen for task-status-updated where status: done. Reference
  • Method B (Polling): Poll GET /v1/collections/{collection_id}/finalize/ until status: done. Reference

• 7. Fetch the Report

  • Action: Retrieve the standardized income or employment data.
  • API Call (Income): GET /v1/links/{link_id}/income/report/
  • API Call (Employment): GET /v1/links/{link_id}/employment/report/
  • Reference: Income Verification