Removing 'um' from a recording is harder than it sounds

A tool called 'erm' was created to remove filler words like 'um' from audio recordings, but it's more complex than expected. The tool uses machine learning to identify filler words and remove them, but it can be tricky to get right. This matters because clear audio recordings are important for many applications, including podcasts and interviews. To use the tool, you can install it locally and use it to strip filler words from your recordings.