# Creating a Template

Creating a template occurs on the top panel of the main menu in the Designer under the "Development" tab — section "Record" — "Extract by Template".

<figure><img src="https://sherparpa.ru/wp-content/uploads/2023/11/image6-624w112h.png" alt=""><figcaption></figcaption></figure>

**Extract by Template**

When the "Extract by Template" button is clicked, the "Template Generator" (Templater) processing opens.

When creating a new template, the document is automatically recognized. Since we are creating a template for a document that we have not worked with before, we need to create new anchors and bindings to the values that will be processed by the robot specifically in this type of document. Therefore, we respond "No" to the pop-up message "Do you want to import all tables from the document into the template?"

<figure><img src="https://sherparpa.ru/wp-content/uploads/2023/11/image7-605w131h.png" alt=""><figcaption></figcaption></figure>

**Text Recognition**

Text recognition in the Templater is performed using built-in and external recognition modules:

1. Tesseract OCR
2. ABBYY Cloud OCR
3. ABBYY Fine Reader
4. Microsoft OCR
5. Yandex Vision

The Sherpa platform integrates several OCR modules. Two of them are offline (shipped with the robot) and will work without an internet connection. These are Tesseract OCR and Microsoft OCR.

Yandex Vision and ABBYY OCR are online modules, meaning they use the functions of the respective cloud modules.

ABBYY Fine Reader is a commercial offline module, and its use requires a separate license.

The platform allows you to configure the script's operation with image recognition and switch between these engines at any time. To do this, you need to select the desired engine from the list.

<figure><img src="https://sherparpa.ru/wp-content/uploads/2023/11/image8-605w207h.png" alt=""><figcaption></figcaption></figure>

If the document being recognized is of high quality, with high resolution, it is sufficient to use free OCR modules.

For processing low-quality documents, it is better to use paid solutions.

After the main settings parameters are configured, the document needs to be processed. The action is initiated by clicking the "Process Document" button.

<figure><img src="https://sherparpa.ru/wp-content/uploads/2023/11/image9-605w321h-2.png" alt=""><figcaption></figcaption></figure>

The processed document looks as shown in the screenshot above.

**Save Template**

To continue working with the template, it needs to be saved by clicking on the icon ![](https://sherparpa.ru/wp-content/uploads/2023/11/image10-24w23h.png) in the upper left corner of the main action panel. Next, specify the path for saving the template and the file name. The file name should be based on the type of document, for example, "Cost Report for Completed Works and Expenses".

<figure><img src="https://sherparpa.ru/wp-content/uploads/2023/11/image11-605w309h.png" alt=""><figcaption></figcaption></figure>
