Return_to_Blueprints
Data Pipeline/Phase_02/15 min build

Enterprise Support RAG Pipeline

A standard Retrieval-Augmented Generation pipeline optimized for enterprise support. It pairs dense text embeddings with Claude 3.5 Sonnet to deliver context-aware, low-latency streaming responses.

Execution_Steps

  1. 01

    Initialize Search Modalities

    Drag an `input` node and connect it directly to an `embedding` node. Open the config for the Embedding node and select "text-embedding-3-large" from OpenAI to ensure high-dimensional semantic capture.

  2. 02

    Configure Vector Retrieval

    Add a `vector_store` node. Connect the Embedding node output to the Vector Store input. In the configuration sheet, set your Provider (e.g., Pinecone/Weaviate), define the Index Name, and set Top K to 5.

  3. 03

    Deploy the Reasoning Engine

    Attach an `llm_call` node to the Vector Store. Select Anthropic as the provider and "claude-3-5-sonnet" as the model. Context injection is handled inherently by passing the vector store payload into the LLM context window. Allocate 3000 tokens for the system prompt to accommodate context.

  4. 04

    Enable Output Terminals

    Connect the LLM Call to a single `output` node. Run Auto Layout [Context Menu -> Auto Layout -> Horizontal] to clean up the graph.

Expected_Metrics

P50_LATENCY:< 1200ms
COST_SAVING:45.0%
SLA_LIMIT:2500ms

Ready to verify?

Open the canvas and simulate these parameters in real-time.

Node_Architecture

inputUser QueryChat Mode
embeddingVector Embeddertext-embedding-3-large
vector_storeKnowledge BaseTop-K: 5
llm_callSupport Agentclaude-3-5-sonnet
outputClient StreamStreaming Output