Skip to content

Pinecone

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

Type: pinecone/v1


Connections

Version: 1

null

Properties

NameLabelTypeControl TypeDescriptionRequired
embeddingApiKeyOpen AI API KeySTRINGTEXTThe API key for the OpenAI API which is used to generate embeddings.true
apiKeyPinecone API KeySTRINGTEXTThe API key for the Pinecone API.true
hostHostSTRINGTEXTUrl of the host.true

Actions

Data Query

Name: dataQuery

Query data from a Pinecone vector store using OpenAI embeddings.

Properties

NameLabelTypeControl TypeDescriptionRequired
queryQuerySTRINGTEXTThe query to be executed.true

Load Data

Name: loadData

Loads data into a Pinecone vector store using OpenAI embeddings.

Properties

NameLabelTypeControl TypeDescriptionRequired
documentTypeDocument TypeSTRINGSELECTThe type of the document.true
jsonKeysToUseJSON Keys to Use[STRING]ARRAY_BUILDERJson keys on which extraction of content is based. If no keys are specified, it uses the entire JSON object as content.false
documentFILE_ENTRYFILE_ENTRY
useTokenTextSplitterUse Token Text SplitterBOOLEANSELECTWhether to use the token text splitter.true
tokenTextSplitterToken Text Splitter{INTEGER(defaultChunkSize), INTEGER(minChunkSizeChars), INTEGER(minChunkLengthToEmbed), INTEGER(maxNumChunks), BOOLEAN(keepSeparator)}OBJECT_BUILDERSplits text into chunks based on token count, using the CL100K_BASE encoding.true
useKeywordEnricherUse Keyword Metadata EnricherBOOLEANSELECTWhether to use the keyword metadata enricher.true
keywordMetadataEnricherKeyword Metadata Enricher{INTEGER(keywordCount)}OBJECT_BUILDERExtract keywords from document content and add them as metadata.true
useSummaryEnricherUse Summary Metadata EnricherBOOLEANSELECTWhether to use the summary enricher.true
summaryMetadataEnricherSummary Metadata Enricher{[STRING](summaryTypes)}OBJECT_BUILDERSummarize the document content and add the summaries as metadata.true