Pinecone

Reference

---

Pinecone is a vector database designed for efficient similarity search and storage of high-dimensional data, commonly used in machine learning and AI applications.

Categories: [artificial-intelligence]

Version: 1

---

Connections

Version: 1

null

Properties

Name	Type	Control Type	Description
Open AI API Key	STRING	TEXT	The API key for the OpenAI API which is used to generate embeddings.
Pinecone API Key	STRING	TEXT	The API key for the Pinecone API.
Environment	STRING	TEXT	Pinecone environment.
Project ID	STRING	TEXT	Pinecone project ID.
Index Name	STRING	TEXT	Pinecone index name.

---

Actions

Data Query

Query data from a Pinecone vector store using OpenAI embeddings.

Properties

Name	Type	Control Type	Description
Query	STRING	TEXT	The query to be executed.

Load Data

Loads data into a Pinecone vector store using OpenAI embeddings.

Properties

Name	Type	Control Type	Description
Document Type	STRING	SELECT	The type of the document.
JSON Keys to Use	[STRING]	ARRAY_BUILDER	Json keys on which extraction of content is based. If no keys are specified, it uses the entire JSON object as content.
FILE_ENTRY	FILE_ENTRY
Use Token Text Splitter	BOOLEAN	SELECT	Whether to use the token text splitter.
Token Text Splitter	{INTEGER(defaultChunkSize), INTEGER(minChunkSizeChars), INTEGER(minChunkLengthToEmbed), INTEGER(maxNumChunks), BOOLEAN(keepSeparator)}	OBJECT_BUILDER	Splits text into chunks based on token count, using the CL100K_BASE encoding.
Use Keyword Metadata Enricher	BOOLEAN	SELECT	Whether to use the keyword metadata enricher.
Keyword Metadata Enricher	{INTEGER(keywordCount)}	OBJECT_BUILDER	Extract keywords from document content and add them as metadata.
Use Summary Metadata Enricher	BOOLEAN	SELECT	Whether to use the summary enricher.
Summary Metadata Enricher	{[STRING](summaryTypes)}	OBJECT_BUILDER	Summarize the document content and add the summaries as metadata.