> ## Content Index
> Fetch the complete content index at: https://blog.jestor.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# How document reading with AI works
- URL: https://blog.jestor.com/document-reading-with-ai-ocr/
- Published: 2026-09-16T18:49:18.000Z
- Updated: 2026-09-16T18:49:18.000Z
- Author: Bruno Bannach

**Document reading with AI**, commonly called **OCR**, turns a photo or file of a document into structured data. In **Jestor**, you attach the document to the OCR field and the record fields are filled automatically, with no typing involved.

## Why manual data entry survives in so many operations

Almost every process starts with a document someone else produced: an invoice, a receipt, a contract, an ID. The data inside it is what the system actually needs.

Historically the only bridge between the two was a person typing. That step is slow, error prone and impossible to scale without adding headcount.

Where manual entry hurts most:

- Invoices retyped into a payables workflow
- Receipts entered one by one into expense reports
- Contract dates and values transcribed from PDFs
- Registration documents copied into onboarding forms
- Typos that only surface when a payment fails

## What OCR can and cannot do for you

| Task                           | Handled by AI reading             | Still needs a person          |
| ------------------------------ | --------------------------------- | ----------------------------- |
| Extract fields from a document | Yes, into the fields you selected | Validating unusual layouts    |
| Create the record              | Yes, filled automatically         | Confirming before it advances |
| Read an attachment on a ticket | Yes, feeding the agent's context  | Deciding on exceptions        |
| Judge commercial terms         | No                                | Legal and business review     |
| Execute the tax entry          | No, that belongs to the ERP       | Accounting team or ERP rules  |

## Step by step to put it to work

1. **Choose the process** where documents arrive most often, usually payables or contracts.
2. **Add an OCR field** to the table that stores those records.
3. **Select which fields** should be filled from the document.
4. **Attach a real document** and review what was extracted.
5. **Add a validation stage** so a person confirms before the record advances.
6. **Connect the automation** that sends the confirmed record onward, for example into the **ERP**.

## Where this fits in the wider operation

- **Payables:** an invoice photo becomes a request ready for approval
- **Contracts:** dates, parties and values extracted into a tracked record
- **Support:** an agent reads the attachment before answering the ticket
- **Onboarding:** registration documents turned into a complete profile

Usage is measurable too, with processed pages and **AI** credit consumption visible inside the platform.

## Frequently Asked Questions

### Does it work with a phone photo of a document?

Yes. Attaching the image to the OCR field triggers the reading. See [jestor.com](https://www.jestor.com/?ref=blog.jestor.com).

### Which fields get filled automatically?

The ones you select when configuring the OCR field for that table.

### Does the AI issue the invoice or post the tax entry?

No. Those executions belong to the ERP. Jestor structures and sends the data.

## Video Tutorial: Step by Step

With **Jestor**, you can automate workflows, connect teams, and build internal systems your way, all without code and powered by **AI**. Discover Jestor at [jestor.com](https://www.jestor.com/?ref=blog.jestor.com) and see how to take your company's operations to a new level of efficiency and control.