> ## Documentation Index
> Fetch the complete documentation index at: https://docs.neum.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Source Connector

> Connector to extract data from data sources

## Overview

<img className="rounded-lg" src="https://mintlify.s3-us-west-1.amazonaws.com/neumai/images/Data_PreProcessing.png" alt="Data PreProcessing" />

The `SourceConnector` is the central unit that takes care of extracting and pre-processing data ahead of embedding. Learn more about data [pre-processing with Neum AI](/local-development/data-preprocessing).

The `SourceConnector` takes three arguements:

* 1 `Data Connector`
* 1 `Loader`
* 1 `Chunker`

Additionally a `custom_metadata` property is allowed. The metadata provided in this field will be added to every `NeumDocument` extracted by the `SourceConnector`.

## Usage

<CodeGroup>
  ```python Local Development
  from neumai.Sources.SourceConnector import SourceConnector

  source_connector =  SourceConnector(
      data_connector = <Data Connector>,
      loader = <Loader>,
      chunker = <Chunker>,
      custom_metadata = {"key":"test"}
  )
  ```

  ```json Cloud
  {
      "sources":[
          {
              "data_connector": {
                  <Add Data Connector configuration>
              },
              "loader": {
                  <Add Loader configuration>
              },
              "chunker": {
                  <Add Chunker configuration>
              },
              "custom_metadata":{
                  "key":"test"
              }
          }
      ]
  }
  ```
</CodeGroup>