Skip to main content
Before classification can begin, you need to tell Pulse Classifier where your response data lives. You can either upload a file from your computer or connect to an existing Smartinterview survey question. Once the data is loaded, you map the columns so the AI knows which text to classify and how to identify each respondent in the output.

​
Supported file formats

Pulse Classifier accepts the following file types:
FormatExtensionNotes
Excel workbook.xlsx, .xlsSupports multiple sheets; one sheet is selected for classification.
SPSS data file.savVariable labels are used as column names where available.
Files with very large numbers of responses may take longer to process and consume more tokens. See token usage for details.

​
Uploading a file

Drag your file from your file manager directly onto the upload area on the Pulse Classifier page. The file uploads immediately and the sheet list loads automatically.
Once the file is uploaded, Pulse Classifier reads the available sheets and presents them in the configuration panel.

​
Configuring the data source

After upload, you map four key fields before proceeding to topic generation.

​
Sheet name

Select the worksheet that contains your response data from the Sheet dropdown. If your file has a sheet named FilesQO, Pulse Classifier selects it automatically. Otherwise, the first sheet is selected by default.

​
Respondent ID column

Select the column that uniquely identifies each respondent from the Respondent ID dropdown. This value is carried through to the export so you can join results back to your original dataset. If your file has no respondent ID column, select None — Pulse Classifier assigns sequential identifiers automatically.

​
Response text column

Select the column that contains the open-ended text responses from the Response column dropdown. Pulse Classifier auto-detects columns whose names include words like “réponses”, “answers”, or “verbatim” and pre-selects the most likely match.

​
Context columns (optional)

If your file includes a question text column (Q1) or a follow-up column (Q1a), you can map them here. These columns are carried into the export but do not affect classification.

​
Multi-column classification

If your file has several open-ended columns that each require their own topic plan, you can set up multiple coding plans — one per column group. Pulse Classifier automatically detects text columns in your file and groups them into an initial plan. Use Add plan to create additional plans with different topic sets and classification parameters.
Assign each coding plan a descriptive name that matches the survey question it covers. This makes the exported output easier to navigate.

​
Using an existing Smartinterview survey question

Instead of uploading a file, you can classify responses collected directly through a Smartinterview survey.
1

Select a survey

Choose a survey from the Select survey panel on the Pulse Classifier page.
2

Select a question

Choose the open-ended question whose responses you want to classify. Only questions with text responses are shown.
3

Proceed to topic generation

Click Continue. Pulse Classifier loads the responses for that question and moves you to the topic generation step.
When classifying a survey question, you do not need to configure a sheet or column — Pulse Classifier fetches the response data directly from the survey.