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

# Batch mode

Batch Mode enables users to upload a CSV file and run an AI agent on each row of the spreadsheet. Each row is treated as a separate user query, allowing for efficient processing of multiple requests.

<Note>
  Batch Mode processes each request independently, including tool use and
  invoking AI teammates, just as it would in the ChatUI.
</Note>

## Use Cases

* Mass data analysis
* Bulk web scraping
* Large-scale content generation
* Automated data enrichment

<Tip>
  Batch Mode works exceptionally well with the web scrape tool for mass scraping
  operations.
</Tip>

## How to Use Batch Mode

Follow these steps to utilize Batch Mode:

<Steps>
  <Step title="Access Batch Mode">
    Navigate to the top right corner of the desired agent and click on the
    "Batch Mode" button.

    <Frame>
      <img src="https://mintcdn.com/invictaai/rLU6J79e0_1ZhRdX/platform-capabilities/batch-mode-icon.png?fit=max&auto=format&n=rLU6J79e0_1ZhRdX&q=85&s=fd54760d9d605f4c17b98040522a6de0" className="rounded-xl" alt="Batch Mode" width="550" height="238" data-path="platform-capabilities/batch-mode-icon.png" />
    </Frame>
  </Step>

  <Step title="Upload CSV">Upload the CSV file you want to process.</Step>

  <Step title="Configure Advanced Options (Optional)">
    In advanced options, you can:

    * Select specific columns for the agent to
      consider
    * Choose the starting and ending rows for processing
  </Step>

  <Step title="Enter Instructions">
    Provide the instruction that will be applied to every row of the
    spreadsheet.
  </Step>

  <Step title="Run the Batch">Execute the batch process.</Step>

  <Step title="Download Results">
    Download the results in either PDF or CSV format: - CSV: The agent's output
    will appear as a new column called "Invicta Output" at the end of the
    spreadsheet. - PDF: Each output will be on a separate page.
  </Step>
</Steps>

## Important Considerations

<Warning>
  Be mindful of your Credits consumption. Batch Mode can quickly use a
  significant amount of resources, especially for large datasets.
</Warning>

<Warning>
  Exercise caution when using tools that may trigger account blocks or hit rate
  limits. For example, mass email sending could be flagged as spam, and some
  integrations may have request limitations.
</Warning>

## Best Practices

1. Start with a small subset of your data to test the process before running a full batch.
2. Monitor your Credits usage to avoid unexpected depletion.
3. Exercise caution when using tools that interact with external systems or services. Be mindful of web scraping practices and ensure compliance with target website terms of service. When sending emails, consider spam regulations and recipient preferences. For API calls, respect rate limits and usage policies of third-party services.
4. Regularly check the output quality to ensure the AI agent is performing as expected across all rows.

<Info>
  {" "}

  When using tools that interact with external systems, always consider the potential
  impact on those systems and your account status with various services.{" "}
</Info>

By leveraging Batch Mode, you can significantly increase the efficiency and scale of your AI agent operations, enabling you to process large volumes of data with ease.