ReadyFlow overview: S3 to Pinecone [Technical Preview]
You can use the S3 to Pinecone [Technical Preview] ReadyFlow to consume PDF documents from S3, vectorize them using an OpenAI model and write the results to Pinecone.
This ReadyFlow consumes PDF documents from a source S3 location, partitions the PDFs, chunks the data, and transforms/splits the data into the required input format for the PutPinecone processor. The PutPinecone processor then vectorizes the data and writes the embeddings to a destination Pinecone index. The default OpenAI model is 'text-embedding-ada-002'. An OpenAI API key and a Pinecone API key are required to run this flow. Define a KPI on the failure_WriteToPinecone connection to monitor failed write operations.
| S3 to Pinecone [Technical Preview] ReadyFlow details | |
|---|---|
| Source | Cloudera managed Amazon S3 |
| Source Format | |
| Destination | Pinecone |
| Destination Format | Vector DB |
