Vector Store Search
A 3-step recipe with OpenAI. We can help you get access.
Recipe · 3 steps
Vector Store Search
with OpenAI
Ingest a file into a vector store and search it once indexing completes.
- Attach the uploaded file to the vector store.
- Poll the vector store file until indexing reaches a terminal status, branching to the search step when completed and looping while in progress.
- Search the vector store for chunks relevant to the query.
What you'll need to have ready
- apiKey: OpenAI API key used as a Bearer token.
- vectorStoreId: The id of an existing vector store to ingest into and search.
- fileId: The id of a previously uploaded file to attach to the vector store.
- query: The natural language query to search the vector store for.
For your developer
Arazzo workflow openai:vector-store-search, 3 operations: createVectorStoreFile → getVectorStoreFile → searchVectorStore. Specs on APIs.io.