Skip to content

About us

Welcome to the Japan group for Neo4j, the world’s leading graph database. AI is only as good as the context you give it, and context is connected data. That’s why graphs sit at the center of modern AI: GraphRAG, knowledge graphs, and context graphs give LLMs and agents the relationships they need to retrieve the right facts, reason across them, and explain their answers.
We host regular events with talks and hands-on sessions on building AI applications with graphs, from GraphRAG and agent memory to context graphs and the core Neo4j database.
Whether you’re new to graphs or an experienced graphista, you’re welcome. Join us afterwards for a drink on us. It’s a great chance to network (see, graphs are everywhere!) and to ask the Neo4j team your questions.

Sponsors

Building Neo4j-Powered Apps with Gen-AI

Building Neo4j-Powered Apps with Gen-AI

A comprehensive guide to building GenAI applications using Neo4j's KGs.

Free Hands-on Online Training

Free Hands-on Online Training

Learn about LLMs + Knowledge Graphs, RAG and more

Neo4j Community Forum

Neo4j Community Forum

Join the Neo4j experts in the forum for Graph Database knowledge & more!

Essential GraphRAG Ebook

Essential GraphRAG Ebook

A comprehensive guide on how to build a GraphRAG system from scratch.

Upcoming events

2

See all
  • Network event
    NODES 2026

    NODES 2026

    ·
    Online
    Online
    56 attendees from 9 groups

    NODES, the biggest community gathering dedicated to engineering better intelligence returns on November 12, 2026, for its eighth year.
    Join thousands of developers and data and AI practitioners for this free, 24-hour
    conference to learn about the latest advancements in intelligent systems and gain insights from speakers who showcase their implementations, tools, and models.

    The agenda is out!!! Here is a snapshot from the AI Engineering track

    • Agent Memory Beyond Vectors: Queryable Context Graphs for Reliable AI Systems, Auxten Wang, ClickHouse
    • Overcoming Temporal Blindness: Serverless Cross-Temporal Reasoning with Neo4j and LangChain, Shuva Jyoti Kar, Palo Alto Networks
    • The Missing Layer: Why AI Memory Should Start with Humans, Not Agents. Doriana Negruț, Luminaria
    • Unlocking Multi-Hop Reasoning with GraphRAG: Divakar Kumar, FlyersSoft
    • Building an AI Customer Support Agent with Mastra & Neo4j: Abhishek Das, SourcingXPress
    • "What Connects to This Pump?"- Agentic, Visual Search over Diagrams, Powered by Neo4j: Hiroki Kokubo, JGC Corporation
    • Public Healthcare: Either a Labyrinth or a Graph, GraphRAG Routing on Neo4j, Giulia Marchese, Geen.ai

    Before NODES 2026: Get hands-on with GraphRAG, agent memory, context graphs, and agentic engineering in our 2-hour workshop series—and share NODES with your fellow devs for a chance to win $200 in cool prizes. Come to NODES with a few new patterns, ideas, and questions ready to dig into.

    • Photo of the user
    1 attendee from this group
  • Agent Harness Build Night

    Agent Harness Build Night

    マリックス恵比寿, Japan, 〒150-0013 Tokyo, Shibuya, Ebisu, 4-chōme−4−6 マリックス恵比寿, Shibuya, JP

    Register: https://luma.com/5fjvo0ps

    AIエージェントの出来は、モデルだけでは決まりません。まず、モデルを動かすループと、呼び出せるツールがあります。さらに、コンテキストや記憶の扱い、どこまで実行を許すか、うまくいったかをどう確かめるかも欠かせません。モデルの周りを固めるこうした仕組みを「ハーネス」と呼びます。エージェントの良し悪しの多くは、ここで決まります。
    ​今夜はこのハーネスをつくります。ゼロから最小限のエージェントループを書いてもいいし、SDKにツールやサブエージェントを足してもかまいません。いま使っているエージェントの評価セットをつくるのもよいでしょう。Claude CodeやCodex CLI、オープンソースのハーネスから設計を借りて、自分のエージェントに組み込んでみるのもおすすめです。
    ​数時間じっくり手を動かし、最後にできたところまでを見せ合います。デモのあとは参加者全員で投票し、いちばん票を集めたプロジェクトを終了前に発表します。
    ​スケジュール

    • ​18:00 — 開場・交流タイム
    • ​18:30 — イントロ・ライトニングトーク「ハーネスの中身」
    • ​19:00 — 開発タイム
    • ​20:45 — デモ(1人3分・任意)
    • ​21:15 — 投票・結果発表
    • ​21:30 — 終了
    • Photo of the user
    2 attendees

Group links

Organizers

Neo4j is a Super Organizer