I have about 6000 documents that can be categorized into 6 categories. Some categories may have upto 1500 docs and some may have less number of documents. I will create multiple indexes(say 6) inside one namespace. I will also add meta data along with each record. We can consider One/Two Page pdf documents. The point I want to get cleared here is that how many pods will be used? Which POD type should I use for my RAG application? Which RAG technique can better serve my application
Hi @zohaibramzan28, and welcome to the Pinecone forums!
Thanks for your question.
While you can certainly use pod-based indexes, and start small while learning more about your workloads, have you considered using Pinecone serverless instead?
Here’s our quickstart guide you can follow.
This way, you don’t need to worry about pre-configuring your workload size at all - you can simply upsert vectors and pay for what you use.
There’s also a free tier you could use to experiment and get started.
Here’s a case-study about one of our customers who ingested 30 million documents into Pinecone, which you may be interested in.
Hope that helps!
Best,
Zack