Neo4j + AWS Strands Agent Demo: Financial Crime Investigation Dataset Exploration

This project demonstrates how to integrate Neo4j with the AWS Strands Agent framework to explore and analyze a financial crime dataset using multi-agent orchestration.

Features

• Agent orchestration using the Strands Agent SDK
• Neo4j graph database for financial crime data
• AWS Bedrock for LLM-powered agent reasoning
• Jupyter notebook for interactive exploration
• Example notebook for agent queries and orchestration

Getting Started

Prerequisites

• Python 3.10+
• The notebook uses a publicly hosted Financial crimes investigation dataset
• AWS account with Bedrock access
• • Neo4j Aura or local Neo4j instance if you want to have your own neo4j deployment
• strands-agents and strands-agents-tools

Setup

1. Clone this repository:

   git clone https://github.com/neo4j-partners/strands-neo4j-agent.git
   cd strands-neo4j-demo

2. Create and activate a Python virtual environment:

   python3 -m venv .venv
   source .venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

4. Copy .env.example to .env and fill in your AWS and Neo4j credentials.

Neo4j MCP Server Integration

This project uses the Neo4j MCP server to enable secure, programmatic access to your Neo4j database from the Strands Agent framework. The MCP server acts as a bridge between the agent tools and your Neo4j instance.

Usage

• Open strands-agent-neo4j.ipynb in Sagemaker Jupyter Notebook or VS Code for an interactive demo.
• Run the notebook cells to:
  • Connect to Neo4j (via MCP server) and AWS
  • Query the financial crime dataset using LLM agents
  • Explore the results

Project Structure

• strands-agent-neo4j.ipynb — Main notebook demo
• .env.example — Example environment variables
• requirements.txt — Python dependencies

License

Apache 2.0

About

This demo is maintained by Neo4j Partners and is intended for educational and prototyping purposes.