Advanced AI Enricher

Here's a tutorial to help you set up the Enrich with AI Phantom:

You have two options:

Use a spreadsheet containing your data

Create a spreadsheet with Google Sheets containing the data you want to enrich. Organize the data in columns and make sure to have clear headers for each column.

Make this spreadsheet public so PhantomBuster can access it.

Copy the spreadsheet URL and paste it into your Phantom's setup.

Use data extracted by another Phantom

You can select the results file of any of your previously launched Phantoms by clicking on the "My Phantoms" tab, then choose the file you want from the dropdown menu.

Choose the columns to enrich
Enter the column names from your spreadsheet that you want to use for enrichment. Include only the necessary columns, as not all data may be relevant for the task. Type each column name on a separate line.

In this step, you'll choose between two Enrichment Modes, either processing your data row by row (Individual Rows) or all at once (Bulk). Select the mode that best aligns with your use-case and the nature of your data. Keep in mind the different capabilities of each mode and how they affect your desired results.

In this step, you'll select or create a prompt that the AI will use to enrich your data. Choose from the list of pre-filled prompts, or craft a custom prompt tailored to your specific requirements. Ensure that the prompt is clear and concise to obtain the most accurate and relevant results from the AI. If you're uncertain about crafting a prompt, start with one of the existing options and feel free to modify it as needed.

Here's an example of a prompt for analyzing scraped reviews:

Analyze the following review and determine the overall sentiment (positive, negative, or neutral). Provide a summary of the main points mentioned in the review. Also translate the review into French. Format the response as a JSON object with keys 'sentiment', 'summary' and 'translation', in this parsable JSON format, as a single flat object: { sentiment: "...", summary: "...", translation: "..." }

PhantomBuster’s Domain Name Finder Automation uses public search engines to find a company’s main website domain. This short tutorial walks you through setting inputs, processing options, scheduling, and retrieving results.

• Matching limitations:

  Returns one main domain (best match) per company. Country & Language settings guide search context but cannot strictly restrict results to specific TLDs (e.g., .fr).

• Input requirements:

  Google Sheets must be public (anyone with the link); CSV URL input is available on paid plans only. No session cookies are required.

• Performance & Safety:

  Processes ~14 domains per minute using 1 slot. Search engines may temporarily block shared IPs if requests are too frequent; if stopped, wait ~15 minutes or use a proxy.

• Free plan limits:

  CSV exports include only the first 10 rows. JSON exports, dynamic CSV download links, and CSV uploads as inputs are unavailable.

1. Provide the company names:

   Choose an input source (My Lists, manual names, a Google Sheet/CSV URL, or My Phantoms), then optionally set Country & Language to guide the search context.

For spreadsheet/CSV inputs, the Phantom reads column A by default. To target a different column, specify its header name in the input settings.

• Configure processing settings:

  Add any domains to ignore, set how many companies to process per launch for spreadsheet/CSV inputs, and optionally rename the results file.

"Number of companies to process per launch" applies only to spreadsheet/CSV inputs (default 100). Renaming the results file between launches will create a new file and restart processing from scratch.

• Select launch frequency:

  Run manually, schedule a one-time run, launch repeatedly, trigger it after another Phantom, or use Advanced scheduling for precise timing.

• Optional: Adjust advanced settings:

  Keep defaults unless you need to fine-tune execution limits, retries, notifications, proxies, webhooks, or file management.

• Launch and retrieve results:

  Click Launch, then open the Results tab to view found domains and download or export your results.

For full details and configuration options, see the tutorial on the help center.