agents.rag.simple.enhanced_v3¶

Enhanced SimpleRAG V3 using Enhanced MultiAgent V3.

This package provides SimpleRAG implementation using the Enhanced MultiAgent V3 pattern with performance tracking, debug support, and adaptive routing capabilities.

Classes:

SimpleRAGV3: Main SimpleRAG implementation with Enhanced MultiAgent V3
RetrieverAgent: Specialized agent for document retrieval
AnswerGeneratorAgent: Specialized agent for answer generation
SimpleRAGState: Enhanced state schema for SimpleRAG pipeline

Examples

Basic usage:

from haive.agents.rag.simple.enhanced_v3 import SimpleRAGV3

rag = SimpleRAGV3.from_documents(
    documents=documents,
    embedding_config=embedding_config,
    performance_mode=True
)

result = await rag.arun("What is machine learning?")

With performance tracking:

rag = SimpleRAGV3(
    name="qa_system",
    vector_store_config=vs_config,
    performance_mode=True,
    debug_mode=True
)

result = await rag.arun("Complex query")

# Monitor performance
analysis = rag.analyze_agent_performance()
print(f"Retriever success rate: {analysis['agents']['retriever']['success_rate']}")

agents.rag.simple.enhanced_v3¶

Submodules¶