agents.rag.multi_agent_rag.advanced_workflows

Advanced RAG Workflows - Graph RAG and Agentic RAG Patterns.

This module implements the most sophisticated RAG architectures including Graph RAG, Agentic routing, speculative execution, and self-routing patterns.

Classes

AgenticGraphRAGAgent	Agentic Graph RAG - combines graph reasoning with agentic routing.
AgenticRAGRouterAgent	Agentic RAG Router - intelligently routes queries to different RAG strategies.
AgenticRAGState	RAG state for agentic routing and planning.
GraphRAGAgent	Graph RAG - uses knowledge graph construction and traversal.
GraphRAGState	RAG state for graph-based approaches.
QueryPlanningAgenticRAGAgent	Query Planning Agentic RAG - creates detailed execution plans.
SelfReflectiveAgenticRAGAgent	Self-Reflective Agentic RAG - continuously reflects on and improves.
SelfRouteRAGAgent	Self-Route RAG - dynamically routes itself to different reasoning.
SpeculativeRAGAgent	Speculative RAG - generates multiple possible answer hypotheses.

Functions

build_custom_graph()

Build custom graph for advanced workflows.

Module Contents

class agents.rag.multi_agent_rag.advanced_workflows.AgenticGraphRAGAgent(**kwargs)

Bases: haive.agents.multi.base.MultiAgent

Agentic Graph RAG - combines graph reasoning with agentic routing. and dynamic planning for complex multi-step reasoning.

build_custom_graph()

Build the custom graph for this multi-agent workflow.

Return type:

Any

class agents.rag.multi_agent_rag.advanced_workflows.AgenticRAGRouterAgent(**kwargs)

Bases: haive.agents.multi.base.MultiAgent

Agentic RAG Router - intelligently routes queries to different RAG strategies. based on query type, complexity, and domain.

build_custom_graph()

Build the custom graph for this multi-agent workflow.

Return type:

Any

class agents.rag.multi_agent_rag.advanced_workflows.AgenticRAGState(/, **data)

Bases: haive.core.schema.prebuilt.rag_state.RAGState

RAG state for agentic routing and planning.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:

data (Any)

class agents.rag.multi_agent_rag.advanced_workflows.GraphRAGAgent(**kwargs)

Bases: haive.agents.multi.base.MultiAgent

Graph RAG - uses knowledge graph construction and traversal. for contextually rich retrieval and reasoning.

build_custom_graph()

Build the custom graph for this multi-agent workflow.

Return type:

Any

class agents.rag.multi_agent_rag.advanced_workflows.GraphRAGState(/, **data)

Bases: haive.core.schema.prebuilt.rag_state.RAGState

RAG state for graph-based approaches.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:

data (Any)

class agents.rag.multi_agent_rag.advanced_workflows.QueryPlanningAgenticRAGAgent(**kwargs)

Bases: haive.agents.multi.base.MultiAgent

Query Planning Agentic RAG - creates detailed execution plans. for complex queries requiring multiple reasoning steps.

build_custom_graph()

Build the custom graph for this multi-agent workflow.

Return type:

Any

class agents.rag.multi_agent_rag.advanced_workflows.SelfReflectiveAgenticRAGAgent(**kwargs)

Bases: haive.agents.multi.base.MultiAgent

Self-Reflective Agentic RAG - continuously reflects on and improves. its own reasoning and retrieval processes.

build_custom_graph()

Build the custom graph for this multi-agent workflow.

Return type:

Any

class agents.rag.multi_agent_rag.advanced_workflows.SelfRouteRAGAgent(**kwargs)

Bases: haive.agents.multi.base.MultiAgent

Self-Route RAG - dynamically routes itself to different reasoning. strategies based on intermediate results and confidence levels.

build_custom_graph()

Build the custom graph for this multi-agent workflow.

Return type:

Any

class agents.rag.multi_agent_rag.advanced_workflows.SpeculativeRAGAgent(**kwargs)

Bases: haive.agents.multi.base.MultiAgent

Speculative RAG - generates multiple possible answer hypotheses. in parallel and validates them against retrieved evidence.

build_custom_graph()

Build the custom graph for this multi-agent workflow.

Return type:

Any

agents.rag.multi_agent_rag.advanced_workflows.build_custom_graph()

Build custom graph for advanced workflows.

This is a utility function for creating custom graphs for advanced RAG workflows in this module.

Returns:

Graph configuration or None for default behavior

Return type:

Any