# Form Features - Magic Fill

## Overview

{% hint style="info" %}
SharinPix **Magic Fill** allows you to automatically populate a **text field** with dynamic content generated from an image.\
Using an AI model, Magic Fill analyzes the image and produces text based on a **user-defined prompt**.

This configuration is available directly in the [**SharinPix Form Template Editor**](https://docs.sharinpix.com/forms/form-elements/sharinpix-form-template-editor).

This article covers the following:

* [How to configure Magic Fill on Text and Text Area Fields](#configure-magic-fill-on-text-and-text-area)
* [Demo of the Magic Fill feature with an Address Extraction Example](#demo-address-extraction-example)
  {% endhint %}

## Getting Started

### Configure Magic Fill on Text and Text Area

Magic Fill can be enabled for any Text or Text Area component inside a SharinPix Form.

1. In the **SharinPix Form Template Editor**, open the **Magic Fill** tab of the element.
2. Check the option **“Magic Fill”**.
3. Provide a **prompt** that instructs the AI on what information to extract or how to interpret the image.

The prompt is essential: it tells the AI what kind of text you expect from the uploaded image.

![](https://2647402409-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FrRD1Xcn9HtKcyfQ9Ghyk%2Fuploads%2F1Vy9W7DxknirzZ3AWWCo%2FDOC%20SF%20-%201920%20x%201080%20\(6\).png?alt=media\&token=775f02f6-9d06-4412-929d-23c083c92e3b)

#### Examples of What Magic Fill Can Do

You can use Magic Fill to:

* Ask the AI to **describe** the image
* Extract **specific details** , such as a serial number, condition, or text written on a label
* Read information from **documents, tags, labels, or packaging**
* Parse structured content such as addresses, dates, or reference numbers

Magic Fill can be configured on a question element of type **Text Area** the same way as shown in the image below

![](https://2647402409-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FrRD1Xcn9HtKcyfQ9Ghyk%2Fuploads%2FTcg45fDGBwkaRreCZNNU%2FDOC%20SF%20-%201920%20x%20480.png?alt=media\&token=c362a18c-3b83-400b-99c8-7bd32b913e2e)

### Demo: Address Extraction Example

This example demonstrates the Magic Fill feature within an **Address Extraction** scenario. When a user uploads an image of an address label or document, the **Address** field is automatically filled with the extracted address information.

![](https://2647402409-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FrRD1Xcn9HtKcyfQ9Ghyk%2Fuploads%2FR73cPWuQQ2QyT0esHshq%2FDOC%20Mobile%20-%201920%20x%201080%20\(2\).png?alt=media\&token=b5f9baca-e502-483a-acba-51582430fd6b)

The example below shows a **Proof of Address** photo which has been taken and is going to be used by the **Address** text field that is not yet populated.

![](https://2647402409-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FrRD1Xcn9HtKcyfQ9Ghyk%2Fuploads%2F7yFPfK1vaRK01yIw6RSM%2FDOC%20SF%20-%201920%20x%201080%20\(1\).png?alt=media\&token=a71df74c-1fa7-4824-9378-1bf4226f92e3)

The diagram below shows the form after the **Proof of Address** has been captured. The **Address** field is automatically filled with the AI-extracted address based on the image and the configured Magic Fill prompt.

![](https://2647402409-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FrRD1Xcn9HtKcyfQ9Ghyk%2Fuploads%2Fm3SVq54rjF6UijhIoUND%2FDOC%20Mobile%20-%201920%20x%201080%20\(1\)%20\(1\).png?alt=media\&token=6a90271d-7ca9-46f3-b8e5-1e105398768c)