LiveNeo4j, Graph Data Science and GraphRAG
Model, query and analyse graphs in Neo4j, then build GraphRAG and agent memory on top
Seven weekends of live classes with a graph data practitioner. You go from your first Cypher query to graph algorithms, knowledge graphs built from documents with an LLM, and a GraphRAG app that answers questions vector search alone gets wrong.
Schedule
Weekends only. Pick the slot that fits your day.
7 weekends from Saturday, 31 October 2026 to Sunday, 13 December 2026. Each class is two hours, live on Google Meet, and recorded.
Morning slot
Small batchSaturdays and Sundays, 10:00 AM to 12:00 PM IST
Times in other countries
- From Sat 31 Oct: Sat + Sun, 12:30 AM to 2:30 AM; From Sat 7 Nov: Fri + Sat, 11:30 PM to 1:30 AM US Eastern
- Sat + Sun, 4:30 AM to 6:30 AM London
- Sat + Sun, 5:30 AM to 7:30 AM Central Europe
- Sat + Sun, 8:30 AM to 10:30 AM Dubai
- Sat + Sun, 12:30 PM to 2:30 PM Singapore
Evening slot
Small batchSaturdays and Sundays, 8:30 PM to 10:30 PM IST
Times in other countries
- From Sat 31 Oct: Sat 11:00 AM to 1:00 PM, Sun from 10:00 AM; From Sat 7 Nov: Sat + Sun, 10:00 AM to 12:00 PM US Eastern
- Sat + Sun, 3:00 PM to 5:00 PM London
- Sat + Sun, 4:00 PM to 6:00 PM Central Europe
- Sat + Sun, 7:00 PM to 9:00 PM Dubai
- Sat + Sun, 11:00 PM to 1:00 AM Singapore
Neither time works? Tell me what would.
Outcomes
What you will be able to do
- 1Turn a business problem into a property graph model, and explain why each fact is a node, a relationship or a property.
- 2Write Cypher from simple MATCH queries to path patterns, aggregation and subqueries, and read a PROFILE plan to make a slow query fast.
- 3Choose the right index or constraint (range, text, full-text, vector, uniqueness) and prove the effect with PROFILE.
- 4Load real data with LOAD CSV, APOC and the Python driver, in safe batches.
- 5Run Graph Data Science algorithms (PageRank, Louvain, Node Similarity, FastRP) and use the results to find influencers, communities and look-alike customers.
- 6Train and evaluate a link prediction pipeline in GDS, and spot data leakage in graph machine learning.
- 7Build a knowledge graph from PDFs with an LLM using the neo4j-graphrag package, and answer questions with vector plus graph retrieval.
- 8Store agent memory and a business ontology in a graph, and explain an agent's answer by walking the graph.
- 9Plan a production setup: clustering, backups, role-based access, monitoring, and the choice between Aura and self-managed Neo4j.
Curriculum
Week by week
Every weekend has the same rhythm: Saturday is the concept and a guided lab, Sunday is a build on your own machine and a review together. Every class ends with a hands-on lab.
Saturday
How a graph database thinks
- Nodes, relationships, labels and properties
- Drawing a model on the whiteboard from a real business question
- MATCH, CREATE, MERGE, WHERE and RETURN
Lab: Run Neo4j in Docker, load a small dataset with LOAD CSV and write your first 20 queries.
Sunday
Build: model an online shop
- Customers, orders, products, devices and addresses as one graph
- Common modelling mistakes: a fact stored as a property that should be a node, and very busy nodes
- Review of everyone's model together
Lab: Load a shop dataset and answer 10 business questions with Cypher.
Saturday
Cypher beyond the basics
- WITH, aggregation, UNWIND and OPTIONAL MATCH
- Variable-length and quantified path patterns, shortest paths
- CALL subqueries and COLLECT subqueries
Lab: 25 graded query exercises on the shop graph.
Sunday
Indexes, constraints and PROFILE
- Range, text, full-text and vector indexes
- Uniqueness, existence and key constraints (which ones need Enterprise Edition)
- EXPLAIN vs PROFILE, rows and db hits, rewriting a slow query
Lab: Take three slow queries, read their plans, fix them and measure the difference.
Saturday
APOC and bulk import
- apoc.load.json to read data from files and APIs
- apoc.periodic.iterate to change millions of nodes in batches
- neo4j-admin database import for a first big load
Lab: Load a public JSON dataset with APOC and batch-update it without running out of memory.
Sunday
Build: a Python pipeline into Neo4j
- The official neo4j Python driver: sessions, managed transactions and retries
- Idempotent loads with MERGE so a rerun never duplicates data
- Review: what breaks when the pipeline runs twice
Lab: Write a small ETL job in Python that loads and updates the graph every hour.
Saturday
Graph algorithms with the GDS library
- Graph projections with gds.graph.project, and memory estimation
- Centrality: degree, PageRank and betweenness
- Community detection: weakly connected components, Louvain and Leiden
Lab: Find the most influential accounts and the rings of linked accounts in a fraud dataset.
Sunday
Similarity and node embeddings
- Node Similarity and K-nearest neighbours
- Node embeddings with FastRP and node2vec
- The graphdatascience Python client
Lab: Build a 'customers like you' recommender from embeddings.
Saturday
Link prediction pipelines
- How a GDS link prediction pipeline is built: features, split, training, evaluation
- Node classification pipelines
- Why graph data leaks between train and test, and how GDS splits relationships to stop it
Lab: Train a model that predicts missing links and read its scores honestly.
Sunday
Build: an end-to-end graph ML project
- Combine embeddings, algorithm scores and a pipeline
- Write the results back to the graph and query them
- Review: where the model is right for the wrong reason
Lab: A fraud or recommendation project you can show in an interview.
Saturday
From documents to a knowledge graph
- Extracting entities and relationships with an LLM, guided by a schema
- SimpleKGPipeline in the neo4j-graphrag Python package
- Entity resolution: when two names are the same thing
- Vector indexes inside Neo4j
Lab: Build a knowledge graph from a folder of PDFs.
Sunday
Retrieval with vectors and the graph
- VectorRetriever, VectorCypherRetriever, HybridRetriever and Text2CypherRetriever
- The GraphRAG class: retriever plus LLM
- Vector-only RAG vs graph context on the same questions
Lab: Run my neo4j-context-lab: 12 business questions answered three ways, graded, and compared.
Saturday
Graphs for AI agents
- Agent memory as a graph: conversations, entities and facts with time
- Ontologies: one definition of 'customer' per team, and which agent may read what
- Neo4j's open-source MCP servers for letting an agent query a graph
- Neo4j Context Engine (announced 19 Oct 2026) is in private preview: we build the same idea from open building blocks
Lab: Save an agent's answers as nodes and explain one answer by walking the graph back.
Sunday
Running Neo4j in production
- Clustering: primaries, secondaries and how writes are agreed
- Backup and restore, role-based access control and monitoring (Enterprise features, shown on the evaluation licence)
- Aura vs self-managed: what you give up and what you gain
- Final project demos and review
Lab: Back up a database, restore it into a new one, and add read-only and admin roles.
Labs and projects
What you build
All labs are our own and run on your laptop. The code is in a labs repo you get before the first class.
Fraud ring finder
Shared devices and addresses link accounts. Louvain and WCC find the rings; PageRank ranks who to check first.
Neo4j · GDS · Python
Look-alike recommender
FastRP embeddings and K-nearest neighbours suggest products from what similar customers bought.
GDS · graphdatascience · Cypher
GraphRAG over your PDFs
An LLM builds the knowledge graph; vector plus graph retrieval answers questions and shows its sources.
neo4j-graphrag · OpenAI or Ollama · Neo4j vector index
Explainable agent memory
Every answer an agent gives is saved as a node linked to what it used, so any answer can be traced.
Neo4j · Cypher · Python
Who it is for
Is this for you?
This is for
- Data and backend engineers who want to add graph skills to their work.
- Data scientists who want graph features and graph machine learning.
- GenAI engineers whose RAG app misses answers that depend on relationships.
- Architects deciding whether a graph database fits their problem.
You need
- Basic Python: functions, lists, dictionaries and installing packages with pip.
- Basic SQL or any query language helps, but is not required.
- A laptop with 8 GB of RAM or more that can run Docker.
- An OpenAI API key is optional: every LLM lab also runs on a free local model with Ollama.
Not for
- Someone who has never written any code.
- Someone looking for a certification exam cram course.
Careers
Roles you can target after this
Graph Data Engineer
Model, load and tune graph databases for production systems.
Knowledge Graph Engineer
Build and govern knowledge graphs and ontologies, often for AI.
Graph Data Scientist
Use graph algorithms and graph machine learning for fraud, recommendations and supply chains.
GenAI / RAG Engineer
Add graph retrieval to RAG apps so answers that need relationships come out right.
Companies that hire for these skills include:
Check each company's careers page for current openings. We have no hiring arrangement with any of them. This programme does not promise a job, a placement or a salary.
Your instructor
Roni Das
18 years of building and running enterprise systems. He writes and teaches the courses on systemdesign.academy: 770 system design lessons and 212 AI engineering lessons, with labs he runs himself.
What is included
Everything you get
14 live classes, 28 hours on Google Meet, with time for your questions in every class.
Every recording, with lifetime access on systemdesign.academy.
The labs repo: every lab and project, runnable on your laptop.
A certificate of completion from systemdesign.academy.
A doubt-clearing group for your batch, between classes.
There is no self-paced Neo4j course on the site yet, so this programme does not include one. You keep every class recording for life.
How paying works
- Full fee: ₹15,000 / $300 in one payment.
- Or reserve your seat for ₹2,000 / $40 now, and pay the remaining ₹13,000 / $260 before the batch starts. The deposit is credited in full.
- Every payment follows our refund and cancellation policy.
Questions
Frequently asked questions
Which Neo4j edition do we use?
Neo4j Community Edition in Docker for most labs, which is free. For the production class, features such as backup, role-based access and clustering need Enterprise Edition, which we run under Neo4j's evaluation licence. Aura Free also works for the query labs.
Do I need to pay for an LLM?
No. The GraphRAG labs run on a free local model with Ollama. If you have an OpenAI API key you can use it instead; the labs show the cost of each run.
Is this the Neo4j Context Engine?
No. Neo4j announced Context Engine on 19 October 2026 and it is in private preview. We teach the open building blocks behind the same idea: ontologies, graph retrieval and agent memory in Neo4j.
When does the batch start, and when are the classes?
The batch starts on Saturday, 31 October 2026. Classes are on Saturdays and Sundays for 7 weekends, two hours each, in the morning slot (10:00 AM to 12:00 PM IST) or the evening slot (8:30 PM to 10:30 PM IST). You choose one slot when you enrol. The last class is on Sunday, 13 December 2026.
What if I miss a class?
Every class is recorded. The recording appears in My live classes on systemdesign.academy after the class, and you keep it for life. Bring your questions to the next class or to the batch group.
How does the seat reservation work?
Pay the reservation deposit to hold your seat in the slot you choose. The deposit is credited in full against the fee, so you pay only the rest. Pay the remaining amount before the batch starts; your reservation email has a link to do it.
Can I get a refund?
Every payment follows our refund and cancellation policy. Please read it before you pay: systemdesign.academy/refund-policy.
How many people are in a batch?
It is a small batch: at most 30 people in each slot, so there is time for your questions.
How do I pay?
In India: UPI, cards or netbanking through Razorpay, in rupees. Everywhere else: PayPal, or a Visa or Mastercard card, in US dollars.
Is a self-paced course included?
There is no self-paced Neo4j course on the site yet, so this programme does not include one. You keep every class recording for life.
Do I get a certificate?
Yes. You get a certificate of completion from systemdesign.academy at the end of the programme.
Where do the classes happen?
On Google Meet. Your joining link and the class calendar reach your email before the first class.
Still unsure? Ask on our WhatsApp group, or book a short call on topmate.io/roni.
Need a different time?
Tell me which days and times work for you
If enough people ask for a similar time, I will open a new slot and email you. This does not hold a seat; a seat is held only by enrolling or reserving.