The goal was to make something that transcribed online videos. After realizing that gcloud didn't support URL's, the goal changed to reading downloaded audio files. Then, I decided that it should simply display what the user said to the screen through a react voice recording library. Some cool text manipulation and data could have been done on it (Like most frequent or similar words).
Unfortunately, my teammate left late and I was on my own. I didn't want to present by myself so I just left. First time using gcloud platform. Just used some of their test scripts. Uses CRA and some express boilerplate. I basically had to come up with a very basic design and figure out the gcloud api.
*Note for later, USE ASYNC/AWAIT CLIENTSIDE this will work. Just need to set up a recorder, and set some code to retrieve the downloaded file after recording.