Skip to content
R&D prototypeDeepLearning.AI × Neo4j

Knowledge Graph Creator

Multi-agent knowledge extraction: schema proposal, relation validation and Neo4j graph construction.

Let agents propose the schema; let validation decide what enters the graph.

01Context

Certification project with DeepLearning.AI and Neo4j.

02Problem

Build a graph from unstructured text without hand-writing the schema.

03Constraints

  • Relations must be validated before insertion

04My role

What I personally built

  • Implementation following the Neo4j agentic KG course

What others owned

Independent project — no team. Everything here is mine.

05Architecture

Architecture explorer

The business flow — what happens, in plain words.

01 / 04

Text — Unstructured documents.

06System

Extraction → schema proposal → relation validation → Neo4j construction.

07Challenges

  • Entity deduplication

08Outcome

Certified: Agentic Knowledge Graph Construction.

09What I learned

Schema-first thinking still matters when agents write the schema.

10Stack

  • Python
  • Neo4j
  • Cypher
  • LLM agents

11Evidence

  • Neo4j — Agentic Knowledge Graph Construction