seems like yes with whisper, is there a

Need to use an ingestion pipeline + docstore + vector store for that

(I would skip the cache from this example)
https://docs.llamaindex.ai/en/stable/examples/ingestion/ingestion_gdrive/?h=googledrive

can i just use a local vector store in storage and still let it work?

In addition, is there an easy way to get timestamps from whisper and add to metadata?

Not easily. You need some layer to track what documents have been inserted.

Also not sure on the whisper thing

hmmmm okay

id probably have to implement my own doc loader and add the timestamp stuff manually to metadata

I know deepgram supports timestamps and notarization but idk about whisper

since documents get broken into nodes, you need some top-level tracking. With a consistent document ID, then you can compare hashes of content.

Hope that makes sense (this is what the example above is doing)

ahhh that makes sense, i guess, technically, i can use the filename as a "unique" identifier in someway

I am trying to throw together a super quick RAG chat for my friends so they can talk to BJJ dvds as the dvds get verbose and annoying, then have the llm point them to the timestamp where the question is answered