Vector Store: Pinecone

Pinecone is a managed, cloud-native vector database designed for fast vector search with low query latency. It provides fully managed indices without infrastructure overhead.

1. Configuration Parameters

To connect to Pinecone, configure the vector_store_params dictionary as follows:

Parameter	Type	Default	Description
`vector_store_type`	`string`	`"PineconeVectorStore"`	Identifier of the Pinecone vector store class.
`api_key`	`string`	`None`	Your Pinecone API key.
`environment`	`string`	`None`	Pinecone cluster environment (e.g. `"us-west1-gcp"`).
`index_name`	`string`	`None`	Target index name created in your Pinecone dashboard.

2. Configuration Example

Here is a Python example configuring a ChatBot with Pinecone:

from aih import ChatBot
 
pinecone_config = {
    "vector_store_type": "PineconeVectorStore",
    "api_key": "your_pinecone_api_key",
    "environment": "gcp-starter",
    "index_name": "aih-chats-index"
}
 
chatbot = ChatBot.docx_chat(
    input_files=["./policy.docx"],
    vector_store_params=pinecone_config
)

Qdrant Azure Cognitive Search