Start prompt
- Write in typescript and nodejs
- implement the service that accepts an audio file
- upload audio to minio bucket
- transcribe it using assemblyai
- using openai summarize all utterances you get from assemblyai to compact the text
- convert the utterances into youtube like chapters: 00:00:00 chapter title
- chapter titles should contain the main point of the utterance
- using openai compact the chapters to highlight the most valuable parts of the conversation
- using ghost admin api create a blog post with the transcription. It should contain the short chapter summary, the full transcription compiled from the utterances and the link to the audio file.