Batch Processing

Batch processing is the handling of multiple documents in a single operation. Instead of uploading or extracting one file at a time, you submit many documents together. DocLD supports batch uploads, bulk extraction, and batch-oriented workflows for high-volume document intelligence use cases.

When to Use Batch Processing

Batch processing improves throughput and reduces per-document overhead compared to one-at-a-time processing.

How Batch Works in DocLD

1. Upload — Send multiple files in one request or use a folder/queue as input
2. Queue — Documents enter a processing queue; each gets a job ID
3. Process — Parsing, chunking, and embedding run per document
4. Complete — Use webhooks or polling to know when each document is ready

For extraction, pass multiple documents with the same schema to extract structured data from all of them. Results are returned per document with confidence scores and citations.

Best Practices

• Match schema to document type — Use prebuilt schemas or form detection for mixed batches
• Monitor rate limits — Large batches may be throttled; use workflows for controlled pacing
• Handle failures — Some documents may fail parsing or extraction; design for partial success
• Use webhooks — Avoid polling; register a webhook for completion events

Related Concepts

Batch processing relies on parsing, extraction, and job tracking. Workflows can automate batch runs on a schedule or when new files arrive.