Neo4j Aura Agents

This repository contains step-by-step instructions on how to use the new Aura Agents functionality. These examples demonstrate how to create intelligent agents that can interact with knowledge graphs to answer domain-specific questions.

Getting Started

Prerequisites

1. Git LFS (Large File Storage): This repository contains large backup files that require Git LFS. Install Git LFS before cloning:

   # Install Git LFS (if not already installed)
   git lfs install
   
   # Clone the repository
   git clone https://github.com/neo4j-product-examples/knowledge-graph-agent.git

   If you've already cloned without Git LFS, you can fetch the large files by running:

   git lfs pull

2. Ensure you have access to Neo4j Aura at https://console-preview.neo4j.io/

Setup Steps

1. Clone this repository (with Git LFS as described above)
2. Follow the specific tutorial for your use case
3. Customize the agents and tools for your specific domain requirements

Agents Available

Legal - Commercial Contract Review Agent

Our Contract Review example demonstrates how to build an intelligent agent for legal professionals. The agent can:

• Identifying high-risk contracts with missing or problematic clauses
• Assessing risk factors and compliance issues across contract portfolios
• Finding contracts with similar clauses or terms for comparative analysis
• Identifying all contracts associated with specific organizations
• Identify key clauses for a given contract

📖 View the complete Contract Review Agent tutorial

Financial Services - Know Your Customer (KYC) Agent

Our comprehensive KYC example shows how to build an intelligent agent for fraud investigation and compliance analysis. The agent can:

• Identify customers involved in suspicious transaction rings
• Detect customers linked to "hot properties" (addresses with many residents)
• Find customers who work for multiple companies (potential bridges)
• Provide detailed customer profiles and risk assessments

📖 View the complete KYC Agent tutorial

About Aura Agent

Neo4j's Aura Agent combine the power of large language models with the structured knowledge in a Neo4j knowledge graph. This enables Aura agents to provide accurate, contextually relevant grounded by your knowledge graph. It helps improve explainability of your agent answers.