AI Agents Hackathon: Memories That Last brings together builders, founders, students, and researchers to create the next generation of intelligent agents.
This hackathon focuses on memory-powered AI, combining:
-
MemVerge’s MemMachine → persistent long-term memory for agents
-
Neo4j Graph Database → deep connected reasoning & relationships
-
LLMs + tools + workflows → to build real, production-ready AI systems
Participants will build agents that think, store context, recall information, reason over graphs, and evolve over time — just like humans.
Whether you're designing personal AI companions, multi-agent workflows, graph-aware copilots, or full enterprise automations, this hackathon gives you everything to build AI that never forgets.
Agenda:
17th Dec
- 10:00 AM – Doors open
- 10:00–10:30 AM – Coffee and Networking
- 10:30–10:40 AM – Opening remarks & logistics overview
- 10:45–11:15 AM – Neo4j workshop, join remotely here
- 11:15–11:45 AM – MemMachine workshop join remotely here
- 12:00- 5:00 PM – Team Formation, Lunch, Hacking
18th Dec
Teams continue hacking remotely
10:00 PM - Submissions close
Requirements
Your submission must include:
1. Working AI Agent DemoA project that uses MemMachine and Neo4j in any of these ways:
-
Persistent agent memory
-
Graph-based reasoning
-
Conversational or tool-using agents
-
Multi-agent workflows
-
Recommendation, search, planning, routing, etc.
Include:
-
Source code
-
Architecture overview (README)
-
Setup instructions
-
Configs for MemMachine & Neo4j
Show:
-
What you built
-
How it works
-
What problems it solves
-
Why memory + graph reasoning matters
Include:
-
Problem
-
Solution
-
Tech stack
-
How MemMachine + Neo4j are used
-
Future improvements and extensions
Prizes
First Prize
Second Prize
Third Prize
Devpost Achievements
Submitting to this hackathon could earn you:
Judges
Gauri Nagavkar
Developer Advocate Lead - MemVerge, Inc
William Lyon
Neo4j
Judging Criteria
-
Innovation & Creativity
How unique, original, or creative is the idea? Does the project push boundaries in how AI agents use memory or graphs? -
Use of MemMachine (Memory Layer)
How well does the project leverage persistent memory? Does the agent meaningfully store, recall, and use long-term context? -
Use of Neo4j (Graph Reasoning)
Does the project use graph relationships, Cypher queries, or connected reasoning in a valuable way? Is Neo4j essential to the solution? -
Demo Quality
Is the video clear, concise, and compelling? Does it show the problem, solution, and the memory + graph features effectively? -
User Experience & Design
Is the interface, workflow, or interaction smooth and intuitive? Does the user understand how to interact with the agent?
Questions? Email the hackathon manager
Tell your friends
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

