Skip to main content

CSV Upload Tool

This is a basic user interface we provide where you can upload a CSV file of text to have it scored and returned as a separate CSV. The CSV upload tool is found in your account dashboard. You can initiate an account by going here.

Below are instructions for each required field.

csv-upload-ui

API Version​

Description:

This item reflects the version of the API that will be used for scoring.

  • Available options by default: v2
  • Available options for legacy users: v1, v2
info

Some features (like Norming Context) are only available in API V2.

Norming Context (Required for API V2)​

Description:

Choose the norming context to apply when calculating scores.
Norming adjusts scores based on reference groups or baselines.

important

This option is only available when using API V2.

Text Column Index (Required)​

Description:

Provide the 1-based index of the column that contains the text samples to be scored in your CSV.

Example:
If your CSV format is like this:

id,text,speaker
1,Hello!,Joe

Then enter: 2 (because "text" is the second column).

Contains Header Row?​

Description:

Indicate whether your CSV file contains a header row (column names).

Options:

  • yes – CSV has headers (e.g., id,text,speaker)

  • no – First row is data

Delimiter Type​

Description:

Select the delimiter used in your CSV file.

Options:

  • comma (default)
  • tab
  • semicolon
  • pipe

Upload CSV File​

Instructions:

Drag and drop your CSV file into the upload area, or click to select it manually.

Limits:

  • Maximum: 25,000 rows/10MB
  • If your word limit is exceeded, an error message will appear
note

If you surpass your word limit while a CSV is being scored, the output will include the rows after the threshold was reached but they will not contain any scores. Additionally, if you submit a file with more than 25,000 rows, only the first 25,000 will be included in the output.

Text Column Preview​

After uploading your CSV and setting the Text Column Index, a preview of the extracted text will appear here for verification.

Submit​

Click the Submit button once all fields are complete and your file is uploaded.

CSV Formatting FAQ​

This covers best practices for structuring your CSVs when uploading text data for scoring or analysis.

What is the simplest valid CSV format?​

The most basic structure includes:

  • A Unique ID column
  • A Text column containing the content you want to analyze

See below for a basic example:

| Unique ID | Text           |
|-----------|----------------|
| 1         | Text sample 1  |
| 2         | Text sample 2  |
| 3         | Text sample 3  |

Each row is treated as one separate sample of language.

Can I include other metadata?​

Yes — if you have additional metadata like:

  • Speaker identity
  • Date
  • Demographics (age, gender, etc.)
  • Annotated labels (job role, conversation type, etc.)

...you can include those as additional columns.

See the two formats below for examples with metadata like:

| Unique ID | Text              | Date       | Exec or Analyst |
|-----------|-------------------|------------|-----------------|
| 1         | Text sample       | 01-Aug-25  | Analyst         |

and

| Unique ID | Text              | Speaker |
|-----------|-------------------|---------|
| 1         | Text sample       | Tom     |

Should each row be one sentence? One conversation?​

That depends on what you want to analyze.

  • If you want conversation-level insights (e.g., full transcripts), each row should be a full conversation.
  • If you want individual-level insights (e.g., per executive or staff member), each row should contain that person's full text.

Each row = one unit of analysis.

Visit this page for more in-depth information about organizing text for different outcomes.

Can I upload multiple files?​

No. The CSV upload tool only supports one document at a time.

Summary of Best Practices​

  • Include a Unique ID column.
  • Organize your Text column according to your intended analysis outcome.
  • Add relevant metadata in extra columns.
  • Use one file for all samples if building a custom norm.