I have a background in linguistics and can identify phonemes by looking at the spectral graph of the audio, allowing me to seamlessly splice together different takes of the sentences if one of the speakers has misspoken. These skills give me a unique edge in a market saturated by automated, AI based workflows which ignore the intricacies of human speech

I take the source audio from the interview

Clean up the background noises such as an AC drone or PC fan noise

I edit the speech as I'm listening to the podcast, including cutting repeated phrases, combining multiple takes by splicing together words

In the second pass, I clean up the rest of the noises like clinking spoons or animal noises