Your first analysis
From a folder of recordings to a report you can read, in about ten minutes.
This walks through a run from start to finish. You need two things ready: an AI provider set up (see Set up Claude or another provider), and a folder holding at least one interview recording or transcript.
1. Put your recordings in a folder
Place your audio, video, subtitle, or transcript files in a single folder. Files for the same person — for example a video and its subtitle — should share a name, so Bristlenose treats them as one session. See Supported files.
2. Run Bristlenose
macOS app
Drag the folder onto the sidebar. Bristlenose creates a project and starts analysing it straight away; progress shows on the project's row.
Command line
Run, replacing interviews with the path to your folder:
bristlenose run interviews
Bristlenose transcribes each recording, identifies who is speaking, extracts quotes, and groups them into sections and themes. The first run downloads a transcription model, so it takes a few minutes; later runs are quicker.
3. Open the report
When the run finishes, the report opens in your browser. The files are written inside your folder, under bristlenose-output/.
4. Read and mark up the quotes
You start on the Quotes page. Each quote is something a participant said. Click the star on a quote to mark it as worth keeping — stars are how you choose which quotes to export later.
Next steps
- Run an analysis — the options available on a run.
- How Bristlenose reads your interviews — what sections, themes, and signals mean.
- Share a report — getting findings out to others.