Current scope
DocsRouter does not provide a Batch API for S3 ingestion, historical-document routing, title verification, seal verification, or an instant title engine.
Deeds and title records can contain typewritten text, handwriting, stamps, and long legal descriptions. /v1/ocr can process supported multi-page documents and return page-level output.
Extract candidate grantor, grantee, parcel, legal-description, and recording fields, then verify them against the source image and authoritative records. OCR output alone cannot establish title, identify every encumbrance, or confirm that a document is official.
For a backlog, build application-controlled queuing, rate limiting, idempotency, and review. DocsRouter does not automatically ingest storage buckets or fan work out across a batch service.