Introduction to Pinecone indexes

Vector Databases for Embeddings with Pinecone

James Chapman

Curriculum Manager, DataCamp

You'll learn to...

Create indexes and ingest vectors
Retrieve and query vectors
Create common AI applications

→ Building and scaling GenAI applications!

semantic search

Indexes

Store vectors
Serve queries and other vector manipulations
Index contains records for each vector, including metadata
Can create multiple indexes

A filing cabinet with four draws, each containing multiple folders.

Pod-Based

A pod containing several indexes.

Choose hardware to create the index → pods
Pod type determines storage, query latency, query throughout

Serverless

Indexes stored in the cloud.

No resource management
Indexes scale automatically
Run on cloud and store in blob
Easier to use and often lower cost

What we'll use in this course

¹ https://docs.pinecone.io/guides/indexes/understanding-indexes

Creating a Pinecone API key

Create a Pinecone Starter account → pinecone.io
Head to "API Keys"
Copy your API key

Pinecone homepage

Pinecone's API keys page

Creating a serverless index

from pinecone import Pinecone, ServerlessSpec


pc = Pinecone(api_key="API_KEY")


pc.create_index(

    name='datacamp-index',

    dimension=1536,

    spec=ServerlessSpec(

        cloud='aws', 
        region='us-east-1'

    )

)

Checking our indexes

pc.list_indexes()

{'indexes': [{'dimension': 1536,
              'host': 'datacamp-index-1eef0f4.svc.aped-4627-b74a.pinecone.io',
              'metric': 'cosine',
              'name': 'datacamp-index',
              'spec': {'serverless': {'cloud': 'aws', 'region': 'us-east-1'}},
              'status': {'ready': True, 'state': 'Ready'}}]}

Let's practice!

Vector Databases for Embeddings with Pinecone