Getting Started with Neo4j

Key Takeaways

Welcome back. In module one, we established what knowledge graphs are, why they matter for enterprise AI, and the landscape of tools and standards. Now, we get our hands dirty. Module two is called getting started with Neo4j, and it covers everything you need to go from a blank laptop to running your first knowledge graph queries. By the end of it, you will have a working Neo4j environment, you will understand Cypher well enough to write your own queries, and you will have built and populated your first graph from scratch. Let us begin. Here is what we will cover. Section 2.1 gets your environment set up. We will look at the two ways to run Neo4j, navigate the browser interface, and connect from Python. Section 2.2 introduces the Cypher query language, the four core clauses, filtering and sorting, and the pattern matching syntax that makes Cypher so expressive. Section 2.3 is hands-on. We will model a small product catalog as a knowledge graph, write create and merge statements, and run our first real queries. And section 2.4 covers data import, loading CSV files, using the APOC library for bulk operations, and connecting Neo4j to existing relational databases. Let us start with getting your environment set up. Section 2.1, setting up your environment. Before we write a single line of Cypher, we need a running Neo4j instance. The good news is that getting up and running takes less than 10 minutes, and it is completely free. Let me show you your options. You have two main ways to run Neo4j. Neo4j Desktop, which runs locally on your machine, and Neo4j Aura, which is the managed cloud service. Neo4j Desktop is a free download for Mac, Windows, and Linux. It gives you full control over your database, no internet connection required. It bundles the Neo4j browser for query editing, Neo4j Bloom for visual exploration, and the APOC and graph data science plugins we will use later in the course. This is what we will use for all hands-on work here. Neo4j Aura is the cloud option. There is a free tier, Aura DB free, that requires no credit card and gives you a hosted Neo4j instance you can access from anywhere in your browser. Automatic backups, updates, and scaling are all handled for you. If you want to use Aura instead of Desktop, everything works identically. The Cypher queries are the same. For this course, we will use Neo4j Desktop. Head to neo4j.com, download it, install it, and create your first project and database. The setup takes about 5 minutes. Once you see the Neo4j browser welcome screen, you are ready. Once Neo4j is running, you have two primary interfaces to know about. Neo4j browser is your main workspace. It is a web-based query editor that you access at localhost:7474 after starting your database from Neo4j Desktop. You type Cypher queries in the top bar and hit control enter or the play button to run them. Results appear either as a visual graph with clickable nodes and edges, or as a table, which is useful for property level data. It also has a query history, saved favorites, and built-in guides. We will spend most of our time in Neo4j browser throughout this course. Neo4j Bloom is a no-code visual graph exploration tool. It is designed for business users and non-technical stakeholders who want to explore a knowledge graph without writing Cypher. You can use natural language search. For example, type "Show me all drugs that treat diabetes." and Bloom translates that into a visual graph. It is a powerful tool for stakeholder demos and for exploring data before you commit to writing queries. For programmatic access, loading data, running queries from your application, or integrating with AI pipelines, you will connect to Neo4j from Python using the official Neo4j driver. The process has three steps. First, install the driver with pip install Neo4j. Second, create a driver instance by passing in the bolt connection URI, which is bolt: //localhost:7687 for a local database, along with your username and password. Third, open a session and use session.run to execute Cypher queries. Look at the code on the right. This is the complete pattern you will use throughout the course. You create a driver once, then open sessions as needed. Each session.run call takes a Cypher string, optionally with parameters, and we will see parameterized queries shortly. The result is an iterable you can loop through. One important note, always close your driver when you are done or use Python's context manager syntax, the with driver.session as s pattern, which handles cleanup automatically.