Agent Framework Integrations

Neo4j Agent Memory provides official integrations for five popular agent frameworks, enabling persistent memory storage backed by Neo4j’s graph database.

Available Integrations

Framework Use Case Status

LangChain

Chains and agents with conversation memory

✅ Production Ready

PydanticAI

Modern type-safe agents with automatic tracing

✅ Production Ready

LlamaIndex

RAG applications with document + graph retrieval

✅ Production Ready

CrewAI

Multi-agent systems with shared memory

✅ Production Ready

OpenAI Agents SDK

OpenAI function calling with persistent memory

✅ Production Ready

Google Cloud

Vertex AI embeddings, ADK agents, MCP server

✅ Production Ready

AWS Strands Agents

AWS Strands SDK with Bedrock and Context Graph tools

✅ Production Ready

Amazon Bedrock Embeddings

Titan and Cohere embeddings via Bedrock

✅ Production Ready

AWS Hybrid Memory

AgentCore + Context Graphs combined

✅ Production Ready

Quick Start

Choose your framework and install:

# LangChain
pip install neo4j-agent-memory[langchain]

# PydanticAI
pip install neo4j-agent-memory[pydantic-ai]

# LlamaIndex
pip install neo4j-agent-memory[llamaindex]

# CrewAI
pip install neo4j-agent-memory[crewai]

# OpenAI Agents SDK
pip install neo4j-agent-memory[openai-agents]

# Google Cloud (Vertex AI + ADK + MCP)
pip install neo4j-agent-memory[google,mcp]

# AWS Strands Agents
pip install neo4j-agent-memory[aws,strands]

# All frameworks
pip install neo4j-agent-memory[all]

Quick Examples

LangChain

from neo4j_agent_memory.integrations.langchain import Neo4jAgentMemory

memory = Neo4jAgentMemory(memory_client=client, session_id="user-123")
context = memory.load_memory_variables({"input": "query"})
memory.save_context({"input": "Hello"}, {"output": "Hi!"})

PydanticAI

from neo4j_agent_memory.integrations.pydantic_ai import MemoryDependency

deps = MemoryDependency(memory_client=client, session_id="user-123")
context = await deps.get_context("query")

LlamaIndex

from neo4j_agent_memory.integrations.llamaindex import Neo4jLlamaIndexMemory

memory = Neo4jLlamaIndexMemory(memory_client=client, session_id="user-123")
nodes = memory.get(input="query")

CrewAI

from neo4j_agent_memory.integrations.crewai import Neo4jCrewMemory

memory = Neo4jCrewMemory(memory_client=client, crew_id="my-crew")
memory.remember("Important fact", metadata={"type": "fact"})

OpenAI Agents SDK

from neo4j_agent_memory.integrations.openai_agents import Neo4jOpenAIMemory

memory = Neo4jOpenAIMemory(memory_client=client, session_id="user-123")
messages = await memory.get_conversation()

Google ADK

from neo4j_agent_memory.integrations.google_adk import Neo4jMemoryService

memory_service = Neo4jMemoryService(memory_client=client, user_id="user-123")
await memory_service.add_session_to_memory(session)
results = await memory_service.search_memories("query")

AWS Strands Agents

from strands import Agent
from neo4j_agent_memory.integrations.strands import context_graph_tools

tools = context_graph_tools(
    neo4j_uri="neo4j+s://...",
    neo4j_password="...",
    embedding_provider="bedrock",
)

agent = Agent(model="anthropic.claude-sonnet-4-20250514-v1:0", tools=tools)

Choosing a Framework

Not sure which integration to use? See the Framework Comparison Guide for a detailed comparison of features, performance, and use cases.

Common Patterns

All integrations share these common capabilities:

Three-Layer Memory

  • Short-Term: Conversation history within a session

  • Long-Term: Entities, preferences, and facts across sessions

  • Reasoning: Task traces and tool usage patterns

Semantic Search

All integrations support semantic search via embeddings:

  • Search messages by similarity

  • Find related entities

  • Discover relevant preferences

Entity Extraction

Automatically extract entities from conversations:

  • People, organizations, locations

  • Events and concepts

  • Custom entity types

Integration Guides