Interactive Playground

Choose your pipeline

Both playgrounds run entirely in your browser — no API keys, no setup. Pick the one that matches your use case.

Generic
RAG Pipeline

The universal RAG flow. Learn how any retrieval-augmented system works — from raw text to a grounded answer.

  • 1
    Ingest Load a document into the pipeline
  • 2
    Chunking Split text into retrievable segments
  • 3
    Embedding Convert chunks to semantic vectors
  • 4
    Search Find the most relevant chunks
  • 5
    Generate Compose a grounded answer
Open RAG Playground →
Salesforce
Data Cloud Pipeline

The Salesforce Data Cloud unstructured RAG flow — from Data Stream ingestion through UDLO, UDMO, and the Search Index.

  • 1
    Data Stream Trigger a source load event
  • 2
    DLO Populate Raw metadata into UDLO fields
  • 3
    DMO Populate UDMO auto-mapped 1:1 from UDLO
  • 4
    Chunking Build Chunk DMO for Search Index
  • 5
    Embedding Generate vectors into IDMO
  • 6
    Retriever Query IDMO and compose answer
  • 7
    Generate Compose grounded answer
Open Data Cloud Playground →