shashankvemuri/meetingAI ? reverse-engineered prompt

Build me a simple web app that lets someone upload or record a meeting audio file and turn it into text.

I want a clean page where I can add an audio file, process it, and then see the transcript right in the browser. It should feel easy to use, with a basic interface for starting the upload, showing that the audio is being handled, and then displaying the written result clearly. If it makes sense, save the uploaded audio in a local folder and keep the app running from a simple Python server on port 4000.

Please make it polished enough to demo, with a nice homepage and a straightforward workflow for someone who just wants their meeting notes in text form. If you need to look up current docs online to make the audio to text part work well, go ahead.