🎙️ Podcast Intelligence
Markus Studio
Real-time podcast transcription powered by local AI. Capture audio from any source, transcribe in multiple languages, and save structured session notes — all running privately on your machine.
Server offline
Launch Studio
Opens the local dashboard at localhost:8080
python3 ~/OpenClaw/podcast-notetaker/markus-studio.py
How It Works
Three steps to live notes
1
Start the server
Run the command above in Terminal. The server loads the Whisper AI model and starts listening on port 8080.
2
Hit Launch Studio
Click the button above or go to localhost:8080. Pick your language and audio device, then press Start Listening.
3
Talk and capture
Speak naturally — transcribed entries appear live with timestamps. Stop when done and your session saves as markdown.
Under the Hood
Built on solid foundations
OpenAI Whisper
State-of-the-art speech recognition running locally
BlackHole 2ch
Virtual audio routing for system audio capture
Revelator IO 24
Direct mic input for podcast recording
Markdown Export
Clean timestamped session files
"A good system doesn't just record — it captures the signal."