The author discusses adding the ability to listen to a webpage with a natural-sounding voice using the Azure speech service.
They use a C# console application to convert MDX markdown files to speech, outputting WAV file bytes. They then optimize cost by converting the WAV bytes to an MP3 file before storing in Azure blob storage, which can be served via a CDN and played in any web browser using the HTML5 audio control.
The author provides code snippets and discusses file formats, tidy up of markdown, and user interface considerations.
















