Introduction

In this guide we'll walk through the process of linking Cerb to Pinecone. You'll be able to use Pinecone's full API in Cerb automations for LLM embeddings.

Get a Pinecone API Key.

Choose API keys in the menu and then click the "Create API Key" button in the top right.

Name the key (eg. cerb) and click Create key.

Copy the API key for use later.

Create the Pinecone service in Cerb

Navigate to Search » Connected Services.
Click the (+) icon in the top right of the list.
Select Pinecone.
Paste the key you copied earlier in the API Key field.
Click the Create button.

Generate embeddings

Pinecone has no agent model record, and it isn't in the Provider list. An agent model exists to be named by a chat automation, and Pinecone generates vector embeddings rather than holding a conversation. So its provider is named inline on the command instead -- there's nothing to create first beyond the connected account above.

start:
  llm.embed/vectorize:
    output: embedding
    inputs:
      llm:
        pinecone:
          model: multilingual-e5-large
      input: Cerb automates helpdesk inboxes and workflows.

The first key under llm: is the provider, so pinecone: is what selects it. Everything nested under that key configures the provider itself.

commands:
  llm.embed:
    allow@bool: yes

Resources

Reference: llm.embed:
Reference: Agent models – for chat providers