Skip to content

getting-transcripts.md

Latest commit

 

History

History
102 lines (71 loc) · 3.97 KB

getting-transcripts.md

File metadata and controls

102 lines (71 loc) · 3.97 KB
Raw

Getting Transcripts

The product-experts skill analyzes full transcripts from Lenny's Podcast. You need to download these transcripts to use the skill.

Download from Dropbox

Lenny Rachitsky generously makes all his podcast transcripts publicly available via Dropbox.

🔗 Download Transcripts

Installation Steps

  1. Visit the Dropbox link above
  2. Select all files (or the ones you want)
    • As of February 2026, there are 303 transcript files in the Dropbox
    • Each file is named after the guest expert (e.g., "Elena Verna.txt")
  3. Download to your transcripts directory
    • Default location: ~/Documents/lennys-podcast-transcripts/
    • The installer creates this directory automatically
  4. Exclude compilation episodes
    • Do not download: Interview Q Compilation.txt
    • This is a compilation episode (not a single expert interview) and creates unusual results
  5. Verify download
    • You should have 302 transcript files (303 minus the compilation)
    • Make sure all files are .txt format
    • Files should be 30-50 pages each (full transcripts, not summaries)

File Format

Each transcript file should:

  • Be in .txt format (plain text)
  • Be named after the expert (e.g., "Ada Chen Rekhi.txt")
  • Contain the full interview transcript
  • Some experts have multiple episodes (e.g., "Elena Verna 4.0.txt")

Important: Do not include compilation episodes like "Interview Q Compilation.txt". The skill is designed for single-expert interviews, and compilation episodes produce inconsistent results.

Verifying Your Download

After downloading, you can verify your setup:

# Check how many transcripts you have
ls ~/Documents/lennys-podcast-transcripts/*.txt | wc -l
# Should show 302 (not 303 - excludes Interview Q Compilation.txt)

# Compare to what's indexed
python3 ~/.claude/skills/product-experts/index/check-index.py

The check script will show if you're missing any transcripts or if you have new ones that aren't indexed yet.

Keeping Transcripts Updated

Lenny publishes new episodes regularly. To stay current:

  1. Check the Dropbox periodically for new files
  2. Download new transcripts to your transcripts directory
  3. Update your index (see How to Use)

Storage Requirements

  • Space needed: ~50-100 MB for 302 transcripts
  • Format: Plain text (.txt files)
  • Location: Any directory on your computer (you specify this in config.json)

Alternative: Self-Transcription

If you have a podcast subscription service that provides transcripts, you can use those instead:

  • Export transcripts as .txt files
  • Name them after the guest expert
  • Place them in your transcripts directory

The skill works with any podcast transcripts in plain text format, though it's optimized for Lenny's Podcast structure.

Supporting Lenny

These transcripts represent hundreds of hours of interviews and are Lenny's intellectual property. While he makes them publicly available, please consider:

This tool helps you get more value from Lenny's incredible archive. Please support him!

Troubleshooting

Problem: Files won't download from Dropbox

  • Try downloading in smaller batches
  • Check your internet connection
  • Make sure you have enough disk space

Problem: Files are in wrong format (PDF, DOCX, etc.)

  • Transcripts must be .txt format
  • If you have other formats, convert them to plain text

Problem: Directory path issues

  • Use absolute paths in config.json
  • Use ~ for your home directory (e.g., ~/Documents/transcripts/)
  • Avoid spaces in directory names, or use quotes if necessary

For more help, see Troubleshooting.