---
schema_version: "1"
id: 123a09f539d212986eba71d8e84f41575e5f760a153c3038cfcd6fca66d72879
title: Configuring Embedding Operations for Einstein AI Connector 2.0
collection: mulesoft
version: current
locale: en-us
portal_url: https://docs.salesforce.com/mulesoft/current/en-us/einstein-ai-connector/latest/configuring-embeddings-operations.md
canonical_url: https://docs.mulesoft.com/einstein-ai-connector/latest/configuring-embeddings-operations
source_site: docs.mulesoft.com
content_type: documentation
content_hash: 54fbc0bc63284b94dddf5240de42472278480361c94d560a5bc50df22ecb51b7
description: MuleSoft Documentation Site
release: "2.0"
latest: true
---
---
product: Einstein AI Connector
version: 2.0
is-latest-version: true
---

# Configuring Embedding Operations for Einstein AI Connector 2.0

> For the full documentation index, see: https://docs.mulesoft.com/llms.txt

## Configure the Adhoc File Query Operation

The **Embedding Adhoc File Query** operation takes a document and ingests it into the vector
database along with its query.

The output of this operation is a set of scores with the complete content of the document that’s the
most likely answer to the query. The vector database is used to identify the numeric representation
of the content before creating the likely score.

1.  Select the operation on the Anypoint Code Builder or Studio canvas.

2.  In the **General** properties tab for the operation, enter these values:

    - **Prompt**

      Plain text prompt to send to the LLM. This field is required.

    - **Input Stream**

      Document content to ingest into the embedding store.

3.  In **Additional properties**, select:

    - **Model API Name**

      Name of the API model that interacts with the LLM.

    - **File Type**

      Type of document to ingest. Values are **PDF**, **TEXT**, and **CSV**.

    - **Option Type**

      How to split the document before embedding. Values are **FULL** and **PARAGRAPH**.

## Configure the Generate From File Operation

The **Embedding generate from file** operation takes a document and ingests it into the vector
database. The output of this operation is a numeric representation of the content.

1.  Select the operation on the Anypoint Code Builder or Studio canvas.

2.  In the **General** properties tab for the operation, enter these values:

    - **Input Stream**

      Document content to ingest into the vector database. This field defaults to `#[payload]`.

3.  In **Additional properties**, select:

    - **Model API Name**

      Name of the API model that interacts with the LLM.

    - **File Type**

      Type of document to ingest. Values are **PDF**, **TEXT**, and **CSV**.

    - **Option Type**

      How to split the document before embedding. Values are **FULL** and **PARAGRAPH**.

### How Data is Parsed

Data from files are parsed in a way that changes the format of the generated embeddings slightly
(the content itself isn’t changed):

- Spaces are removed from the beginning and end of file content, for example:

  Before parsing:

  ```txt
  "

  Para 1
  Para 2



  "
  ```

  After parsing:

  ```txt
  "Para 1
  Para 2"
  ```

- Extra lines between paragraphs are removed, for example:

  Before parsing:

  ```txt
  "
  Para 1


  Para2
  "
  ```

  After parsing:

  ```txt
  "Para 1
  Para 2"
  ```

- The connector provides filtering logic that removes some characters from the generated embedding,
  for example, non-breaking spaces (<0xa0>). These characters are removed to drive more accurate
  embeddings:

<table><colgroup><col> <col> <col> <col> <col></colgroup><thead><tr><th>Character</th><th>Hex Code</th><th>Decimal Code</th><th>Name</th><th>Description</th></tr></thead><tbody><tr><td><p>\u0000</p></td><td><p>0x00</p></td><td><p>0</p></td><td><p>Null (NULL)</p></td><td><p>Marks the end of a string in C-like languages</p></td></tr><tr><td><p>\u0001</p></td><td><p>0x01</p></td><td><p>1</p></td><td><p>Start of Heading (SOH)</p></td><td><p>Used to mark the beginning of a heading in data streams</p></td></tr><tr><td><p>\u0007</p></td><td><p>0x07</p></td><td><p>7</p></td><td><p>Bell (BEL)</p></td><td><p>Triggers a beep or alert sound</p></td></tr><tr><td><p>\u0008</p></td><td><p>0x08</p></td><td><p>8</p></td><td><p>Backspace (BS)</p></td><td><p>Moves the cursor one position backward</p></td></tr><tr><td><p>\u0009</p></td><td><p>0x09</p></td><td><p>9</p></td><td><p>Horizontal Tab (TAB)</p></td><td><p>Inserts a tab space</p></td></tr><tr><td><p>\u000A</p></td><td><p>0x0A</p></td><td><p>10</p></td><td><p>Line Feed (LF)</p></td><td><p>Moves to a new line (Unix newline)</p></td></tr><tr><td><p>\u000D</p></td><td><p>0x0D</p></td><td><p>13</p></td><td><p>Carriage Return (CR)</p></td><td><p>Returns to the beginning of a new line</p></td></tr><tr><td><p>\u001B</p></td><td><p>0x1B</p></td><td><p>27</p></td><td><p>Escape (ESC)</p></td><td><p>Used to introduce escape sequences for control</p></td></tr><tr><td><p>\u001F</p></td><td><p>0x1F</p></td><td><p>31</p></td><td><p>Unit Separator (US)</p></td><td><p>Separates units of information</p></td></tr></tbody></table>

## Configure the Generate From Text Operation

The **Embedding generate from text** operation takes text and ingests it into the vector database.
The output of this operation is a numeric representation of the content.

1.  Select the operation on the Anypoint Code Builder or Studio canvas.

2.  In the **General** properties tab for the operation, enter these values:

    - **Text**

      Text to ingest into the vector database

3.  In **Additional Properties**, select:

    - **Model API Name**

      Name of the API model that interacts with the LLM.

## See Also

- [Using Anypoint Code Builder to Configure a Connector](https://docs.mulesoft.com/connectors/introduction/intro-config-use-acb)

- [Using Studio to Configure a Connector](https://docs.mulesoft.com/connectors/introduction/intro-config-use-studio)

- [Einstein AI Connector Overview](https://docs.mulesoft.com/einstein-ai-connector/latest/)

- [Einstein AI Connector Reference](https://docs.mulesoft.com/einstein-ai-connector/latest/einstein-ai-connector-reference)

- [Salesforce Help](https://help.salesforce.com)