In this article, I will describe the Audio Transcription Workflow (ATW) – a way to use Airtable to make the transcript of your entire media playlist searchable. This is reasonably easy and also quite inexpensive because of recent advancements in speech recognition, and the Turboscribe service which has turned it into an excellent product.
You can use the ATW for any video host and even other types of content such as podcasts and online courses.
Here is the step-by-step process:
Step | Task | Manual Process? | Can use Python Script to automate? |
1 | Add the list of audio or video files into Airtable | ✅ | ✅ Possible to use Python script but also depends on the channel. For example it is quite easy with YouTube which provides an API, a bit more challenging with Rumble which does not provide an API |
2 | Download the audio to your local machine | ✅ | ✅ |
3 | Transcribe the audio using TurboScribe | ✅ | ❌ |
4 | Update speaker labels inside TurboScribe | ✅ | ❌ |
5 | Spell check transcript inside TurboScribe | ✅ | ❌ |
6 | Upload the CSV file to Airtable | ✅ | Doable, but not a good option |
7 | Generate markdown from the CSV file | ❌ This cannot be done manually. It can only be done using a Python script. | ✅ |
8 | Post the article | ✅ | Doable, but not a good option |
9 | Generate search index | ❌ This cannot be done manually. It can only be done using a Python script. | ✅ |