Package ai.pipestream.connector.intake.v1

Interface ConnectorIntakeServiceGrpc.AsyncService

All Known Implementing Classes:
ConnectorIntakeServiceGrpc.ConnectorIntakeServiceImplBase
Enclosing class:
ConnectorIntakeServiceGrpc
public static interface ConnectorIntakeServiceGrpc.AsyncService
 High-throughput document ingestion service for all connector types.
 This service provides the main entry point for document ingestion into the AI Pipestream platform.
 It handles authentication, account verification, metadata enrichment, and routing to the repository layer.
 Supports both structured PipeDoc uploads and raw blob uploads with automatic wrapping.

  • Method Details

    • uploadPipeDoc

      default void uploadPipeDoc(UploadPipeDocRequest request, io.grpc.stub.StreamObserver<UploadPipeDocResponse> responseObserver)
       1. The "Power User" path: Client sends a full PipeDoc
 Best for internal services, advanced connectors, or "Box.com" style integrations
 where the client already knows the structure.

    • uploadBlob

      default void uploadBlob(UploadBlobRequest request, io.grpc.stub.StreamObserver<UploadBlobResponse> responseObserver)
       2. The "Simple" path: Client sends raw bytes (Blob) + minimal metadata
 Best for "I just have a file" scenarios. Intake constructs the PipeDoc wrapper.

    • startCrawlSession

      default void startCrawlSession(StartCrawlSessionRequest request, io.grpc.stub.StreamObserver<StartCrawlSessionResponse> responseObserver)
       Register a new crawl session (Async/Background only)

    • endCrawlSession

      default void endCrawlSession(EndCrawlSessionRequest request, io.grpc.stub.StreamObserver<EndCrawlSessionResponse> responseObserver)
       End a crawl session (for cleanup, orphan detection)

    • heartbeat

      default void heartbeat(HeartbeatRequest request, io.grpc.stub.StreamObserver<HeartbeatResponse> responseObserver)
       Heartbeat to keep session alive