Workflow & integrations

Yes. We offer bulk pricing, enterprise plans, and custom integrations. For large projects, we provide full managed processing, where our team handles the entire workflow from start to finish and tailors the process to your exact document type - often delivering higher-quality results than standard processing.

Yes. In the dashboard, you can select multiple processed documents and download them as a single merged output. This is useful for assembling multi-part scans or grouping related pages into a unified file.

Handwriting OCR doesn't accept uploads directly via email — there's no submit@ address that turns attachments into transcriptions. But you can build that pipeline yourself with very little code using a workflow tool.

Email-to-OCR pattern

The standard recipe:

1. Trigger — incoming email to a dedicated address (Gmail, Outlook, or a custom inbox supported by your workflow tool).
2. Filter — only forward emails matching specific criteria (subject line, sender, attachment type).
3. Forward attachments — extract PDF/image attachments and POST them to the Handwriting OCR API with your token.
4. Webhook back — when the OCR job completes, our webhook fires and your workflow can email the transcription back, drop it into a Drive folder, or push it to your CRM.

Tools that work well

• Make.com — visual workflow builder, free tier covers most light volumes.
• Zapier — easiest for simple linear pipelines.
• Power Automate — native Microsoft integration, good for Outlook-based workflows.

For higher-volume or more complex routing (per-sender folders, conditional re-routing, retry on failure), a small custom script using one of our client libraries or SDKs gives you more control.

If you'd like a starter Make.com or Zapier scenario template for the email-to-OCR pipeline, let us know and we'll share one.

Yes. Custom Extractors let you define fields and automatically extract structured data from handwritten or printed forms. Outputs are available in CSV, XLSX, and JSON formats. They’re ideal for processing invoices, order forms, surveys, and other repeatable layouts.

Yes. The API allows you to process documents without using the dashboard at all. You can integrate OCR directly into your application, automate workflows, or run batch jobs without manual interaction.

Yes — both DOCX and PDF exports support embedding the original page images alongside the transcribed text. This is the safest export option when you need to review the OCR result against the source, or archive both together for compliance.

How to enable it

When exporting from the dashboard, choose the export format (DOCX or PDF) and toggle the option to include the original images. Via the API, set the equivalent flag on your export request — see the API documentation for the exact parameter name.

Things to be aware of

• File size — embedded images make the output significantly larger. A 50-page document that fits in a 100 KB plain DOCX can become tens of megabytes with images embedded.
• Image quality — the embedded image is the page Handwriting OCR processed; it carries the same resolution and compression as the original upload.
• Layout — the transcription typically appears below or beside each page image, not as an invisible text layer. If you need a searchable PDF where the transcription is overlaid as hidden text, let us know — we can scope this for enterprise integrations.

For plain transcriptions you'll edit further, leave the original images turned off — exports stay small and easy to work with.

Yes - all users can access our API. Ideal for automating bulk workflows or integrating OCR into your own software. Our API is fully documented here.

We support automation through the API, which can be integrated with most workflow tools and platforms. Direct processing via email or platforms like Make.com is not supported yet, but many users create automations around the API endpoints to achieve the same effect.

Yes. You can choose the 'tables' extraction option to detect and extract tabular data directly into structured formats like XLSX and JSON. This works for both handwritten and printed tables.

Yes. Your results remain available until the auto-deletion period you choose expires. You can download them again at any time before that point. You can also manually delete documents whenever you like.

Custom extractors turn handwritten or printed forms into structured data — name, date, total, signed-by, etc. — instead of free-form text. They're available on the Pro and Business plans.

Setting one up

The fastest path uses our AI-assisted setup — upload a sample document and the AI examines it and suggests fields automatically. You can accept the suggestions, modify the schema, or go straight to testing.

1. Open Extractors in the dashboard and click New extractor.
2. Name and describe the extractor — a short description helps the model understand the document type.
3. Upload a sample of the form or template. The AI scans it and proposes fields and types.
4. Review the suggested fields (or define your own from scratch):
   • Field names must use letters, numbers, and underscores — no spaces or special characters.
   • Text is the safe default field type — it accepts anything and avoids false validation errors. Switch to a stricter type only when you've confirmed the source is consistent.
   • Add a short description to each field if you want extra steering.
5. Test it — results come back in 1–2 minutes. Refine field names or descriptions if the output isn't what you expected.

Page limits

By default, custom extractors are capped at 10 pages per file. If your documents have a consistent page length — for example, a 2-page form repeated across a PDF — we can lift the limit on request. If your documents have variable page lengths, the limit stays in place to protect accuracy.

Even without the limit, we recommend processing no more than 100 pages per file for best results — split larger jobs into batches.

Using extractors via the API

Once your extractor is saved, you can call it from the API using the extractor action and the extractor's ID. Output is available as XLSX, CSV, or JSON — see the API documentation for the exact request format.

You can upload documents directly through the dashboard or send them through the API. The dashboard is best for one-off or small batches, while the API is ideal for automating larger workflows or integrating Handwriting OCR into your own systems.

Check whether the source PDF or image contains any corrupted or blank pages. If the issue persists, try re-uploading the document or splitting it into smaller files. If you run into repeated issues, upload a few sample pages and contact support—we can quickly determine what’s causing the problem.

For high volumes, the API is the fastest and most efficient method. It allows automated uploads, status checks, and result retrieval without manual steps. If you'd like to completely offload the process, our managed processing service handles everything for you—from ingestion to result delivery.