Data sources

Everything you analyse in Skimle enters your project as a data source. A data source is a collection of files or survey responses that you add and analyse together, and each one carries its own settings, so you can, for example, anonymise interview transcripts while leaving a set of public reports untouched.

Open the Data sources section from the project navigation to see every data source in the project, add a new one, and confirm files you have uploaded.

Types of data source

Click New data source and choose the type that matches your material:

• Text documents: upload PDF, DOCX, RTF, or plain-text files. See supported file formats.
• Audio or video: upload recordings and Skimle transcribes them into documents with speaker labels, then analyses them in place. No separate download step is needed.
• Tabular data: upload a CSV or Excel spreadsheet where each row becomes a document. See CSV import.
• Skimle Ask survey: run an AI-conducted interview; each respondent becomes a document.
• Connect to a data source: pull data directly from an external service (coming soon).

You can also drag files straight onto the project or the home screen: Skimle infers the right data source type from the files you drop.

Uploading and confirming files

When you add files to an upload data source, they first appear as pending. At this stage no credits are used, so you can review what you are about to analyse.

To commit them, select the files and click Confirm. Confirming spends credits based on the volume of your data (see pricing for how credits work). Confirmed files then move through any processing the data source needs (transcription, review, or anonymisation) before they are ready to analyse.

Data source settings

Each data source shows a small set of switches on its confirmation card. They apply to the documents in that data source:

• Automatically analyse: when on, confirmed documents are analysed immediately once all processing (transcription, review, anonymisation) is complete. This setting is shared across every data source in the project. Turn it off if you want to control which documents to include before running an analysis manually. You need at least one analysis for it to take effect.
• Review transcripts is available on Audio or video data sources. When on, you can manually review and correct the transcripts before analysis. See uploading voice files.
• Anonymise documents controls how Skimle handles personally identifiable information. See anonymising documents.

Data sources that carry review or anonymisation settings show a small Review, Anonymise, or Automatic badge on their card so you can see at a glance how each one is configured.

Adding more than one data source

A project can hold several data sources, and you can mix types. Interview transcripts, a spreadsheet of survey responses, and a set of reports can all sit in the same project. If you add a second data source of a type you already have, Skimle offers to either reuse the existing one or create a new one with its own settings.

Once your documents are confirmed and processed, head to the Analyses section to code them into themes and categories.