Interviewing for Your Project: How AI Transcribes and Codes Themes

For oral history projects or capstone interviews, transcription used to take hours. Now Otter does it live, and Claude can pull the patterns.

Some examples

• Otter.ai free tier: 300 minutes/month of live transcription
• OpenAI Whisper API: $0.006/min, more accurate for noisy audio
• Prompt Claude: 'Identify the 5 main themes across these 4 transcripts'
• Ask AI to find quotes that contradict your hypothesis

Try it!

For your next project interview, record (with consent) into Otter. Drop the transcript into Claude with 'Find 5 quote-worthy moments.'

A 30-minute interview used to take 2-3 hours to transcribe by hand. Now Otter.ai, Whisper (free, local), and Apple's built-in Voice Memos transcribe a recording in minutes — usually 95%+ accurate. The transcription is just the start: 'coding' (tagging themes) is where the real research happens.

Some examples

• Otter.ai's free tier gives you 300 minutes/month of transcription with speaker separation — enough for most school projects.
• Whisper (open-source, free, runs on your laptop) is the most accurate transcriber available and works offline — great for sensitive interviews.
• After transcription, prompt Claude: 'Read this interview transcript and tag every passage with one of these themes: [list]. Output a table with theme, quote, and timestamp.'
• Always get verbal recorded consent at the start of the interview ('I'm recording — is that okay?') and offer to share the transcript afterward.

Try it!

Interview a family member for 10 minutes about something they remember — a specific event, their first job, their childhood neighborhood. Record on your phone, upload to Otter.ai, and read the transcript. You just made a primary source that didn't exist before.