the challenge: "Context, connected"

Join the Relate by Neo4j Hackathon, a month-long online event celebrating the launch of Relate by Neo4j at NODES 2026.

Build a project that shows how connected data helps agents, apps, or people find and use the relationships that matter. Your project can be new or extend something you already have. Share your code and documentation so others can learn from and reuse your work.

Relate is an open-source embedded database for AI agents and the engineers who build them. It runs as a library inside your application, with no signup or server to deploy.

  • Hack as an individual.
  • You can work with teammates but please note: Prizes are awarded per winning entry, and not per individual team member.
  • Neo4j employees may participate but are not eligible for the three prizes.

about relate by neo4J

The Resources tab above will get you started quickly

Relate is an embedded database that lets AI agents build and understand relationships across any data. It runs inside your application. Use it from the command line, as a library in Rust or Python, or in a browser tab with WebAssembly. One command installs a skill into the coding agents you already use, so they can find it and start using it on their own. It also stays connected to the Neo4j ecosystem through Cypher, Bolt, HTTP, and MCP.

  • Install and go. Add Relate to your coding agent with one command. No server to run.

  • Query the graph in process. Traverse relationships and run vector searches without network calls.

  • Get agent-ready results. Return JSON with row limits and errors that explain how to fix them.

  • Share graph state safely. Transactions let multiple agents update the same graph without clobbering each other.

  • Keep it lightweight. Store each graph in one file, set memory limits, and delete it when you’re done.

  • Build locally, connect when needed. Pull a subgraph from Neo4j and query it with Cypher through MCP, Bolt, or HTTP.

  • Explore in the browser. Use the WebAssembly build to create interactive graph views, with nothing to install.

🚀GET STARTED

Everything you need to go from zero to your first graph:

🌐  Website: relatedb.ai

⚡  Quickstart and installation: relatedb.ai/docs/get-started/install

💻  GitHub: github.com/relatedb/relate

💡  Relate Ideas: github.com/jexp/relate-ideas

Your idea doesn't have to fit a category. New projects and extensions of existing ones are both welcome.

Requirements

SUBMITTING YOUR PROJECT

Submit directly on Devpost. Judges should be able to clone it, run it, and see your graph without asking you anything.

Required: your public GitHub repo must be tagged with the topic relate. See "Required GitHub topic tag" below.

SUBMISSION CHECKLIST

☐  Project description: the use case, and what connects to what.

☐  Public GitHub repo with a README on how to set up and use it.

☐  GitHub topic tag "relate" added to your repo (required). Adding "relate-neo4j-hackathon" is recommended.

☐  Persistent Relate store: a directory or single file judges can inspect.

☐  Working code (or the prompts you used) with Relate actually running.

☐  Demo video, 1-2 minutes. A screen share or presentation is fine.

REQUIRED GITHUB TOPIC TAG

Your public GitHub repository must include the topic relate. This is how we find and verify entries, and submissions without it may be considered incomplete.

  1. Open your repository on GitHub.
  2. Click the gear icon next to About on the right side of the repo page.
  3. Under Topics, type relate and press Enter.
  4. Optional but recommended: also add relate-neo4j-hackathon.
  5. Click Save changes. Your repo should now appear at github.com/topics/relate.

Already submitted? Add the tag to your repo before the submission deadline. You don't need to resubmit on Devpost.

README TEMPLATE

Your README must answer one question directly: how are you using Relate, and why? Relate can run in memory, as a directory, or as a single file, so tell us which mode you chose and why.

# Project name
One-line pitch.

## Use case
What problem it solves and which relationships matter.

## How it uses Relate
- Mode: in memory | directory | single file
- Why this mode:
- What the graph models:

## Setup and run
Prerequisites, install steps, and the commands to reach a first result.

## Where the data lives
Path to the persistent store in this repo.

GOOD TO KNOW

  • Tag your repo. The relate GitHub topic is required, and your repo must be public so we can verify it.
  • Persist your graph. In-memory data disappears when the app stops, so judges can't inspect it. Include the store in your repo or show it in your demo.
  • Show it running. Entries that are only prompts or AI-generated text, with no running Relate integration, won't be considered.

Hackathon Sponsors

Prizes

3 non-cash prizes
First place
1 winner

Mac Studio (M5 Max, 64GB / 1TB) valued at $3,799

Second place
1 winner

Mac mini (M6, 32GB / 512GB) valued at $1,499

Third place
1 winner

Mac mini (M6, 16GB / 256GB) valued at $899

Judges

Michael Hunger

Michael Hunger
VP Product Innovation at Neo4j

Andreas ("ABK") Kollegger

Andreas ("ABK") Kollegger
Director, Developer Advocacy at Neo4j

William Lyon

William Lyon
Senior Product Manager, AI Innovation at Neo4j

Oskar Hane

Oskar Hane
Senior Software Engineer at Neo4j

Yaru Lin

Yaru Lin
Lead Product Manager @ Neo4j

Kate Smith

Kate Smith
Developer Marketing Manager, AI Capabilities at Neo4j

Jennifer Reif

Jennifer Reif
Senior Developer Advocate at Neo4j

Stephen Chin

Stephen Chin
VP, Developer Relations, at Neo4j

Judging Criteria

  • Creativity: 10 points
    How fresh is the project idea or its use of Relate? Higher scores go to projects that present an original idea or an inventive way to use Relate.
  • Relate and connected context: 30 points
    How effectively does the project use Relate’s graph relationships to support its main use case? This includes how easily judges can use the submission of instructions, running, and exploring with Relate working as described.
  • Functionality and completeness: 25 points
    How much of the project’s core use case works?
  • Usefulness: 20 points
    How clearly does the project address a real problem for an intended user? Who benefits and the does the project clearly show this.
  • Documentation and reproducibility: 15 points
    How clearly does the README explain the project and how to try it?

Questions? Email the hackathon manager

Invite others to compete

Hackathon sponsors