← All courses

🕸️ Neo4j

When the relationship is the data — the property graph, Cypher patterns that look like the picture they match, variable-length paths, and the algorithms that find communities and rank importance. With an episode that tells you to write the question in SQL first.

No lessons published yet — this is the roadmap, so you can see exactly what the course will cover before it lands.

Before you start: NoSQL Concepts — Series 1, and Series 2. The graph shape and the question it answers cheaply are introduced there for every engine, alongside the honest comparison with a recursive query in SQL. This track assumes that comparison has been made and gets on with the modelling.

Before you start: Learn SQL — Series 2, and Series 3. Series 3 · 00 there is the recursive CTE, and it is the thing this product must be measured against. A reader who cannot walk a hierarchy in SQL cannot judge when a graph database earns its place.

Short on time? The 14 episodes marked core are the shortest path to being productive at work. Start there and pick up the rest whenever a job asks for them.

Series 1 · Nodes, Relationships and Cypher

the property graph, and a query language shaped like a drawing

  1. 01

    Nodes, Relationships and Properties — Where the Join Table Went

    coreintermediatecoming soon
  2. 02

    MATCH — a Query That Looks Like the Picture It Is Asking For

    coreintermediatecoming soon
  3. 03

    CREATE, MERGE and DETACH DELETE — and the MERGE That Duplicates Your Nodes

    coreintermediatecoming soon
  4. 04

    Variable-Length Paths — the Question SQL Expresses Badly, and Its Cliff

    coreintermediatecoming soon
  5. 05

    Project — Answer Three Questions That Are All About Paths

    coreintermediate#projectcoming soon

Series 2 · Modelling a Graph

what becomes a node, and proving the model with a query plan

  1. 01

    Node, Property or Relationship — Three Tests, and the Supernode to Avoid

    coreadvancedcoming soon
  2. 02

    From a Relational Schema — What Transcribes Mechanically, and What Must Not

    coreadvancedcoming soon
  3. 03

    Indexes Find the Starting Node — and Then Stop Mattering

    coreadvancedcoming soon
  4. 04

    PROFILE and db hits — Fix the Start of the Plan Before Anything Else

    coreadvancedcoming soon
  5. 05

    Project — Model a Domain, Then Prove Each Query Starts From an Index

    coreadvanced#projectcoming soon

Series 3 · Algorithms, Import, and When Not To

what the graph can compute, and the episode that says no

  1. 01

    Graph Algorithms — Who Matters, Who Clusters, What Connects and What Is Similar

    coreadvancedcoming soon
  2. 02

    Recommendations as a Traversal — a Strong Baseline, and Where It Stops

    advancedcoming soon
  3. 03

    LOAD CSV and admin import — Constraints First, Then Nodes, Then Relationships

    coreadvancedcoming soon
  4. 04

    When Not a Graph — Write the Question Both Ways and See

    coreadvancedcoming soon
  5. 05

    Project — Take a Real Domain and Argue Both Sides Before Choosing

    coreadvanced#projectcoming soon

Not here, on purpose

  • Why a graph at all, and how it differs from the other shapes — covered in NoSQL Concepts, Series 2 · 20, which this course assumes
  • Walking a hierarchy in SQL — covered in Learn SQL, Series 3 · 10, which is the recursive CTE. In this course, Series 3 · 15 compares the two honestly rather than assuming this one wins
  • Graph theory as mathematics — not covered. Series 3 · 00 uses the algorithms and explains what each one answers; the proofs are somebody else's course
  • Running the server — clustering, backups, upgrades, capacity — covered in Data Ops for the discipline
  • Knowledge graphs for retrieval-augmented generation — covered in AI for Data, Series 2, which owns embeddings and retrieval. In this course, Series 3 · 05 is recommendation from the graph itself, with no model involved