Neo4j × hackFront India Agent Memory Build Sprint Pune
A hands on build experience where you learn how Agent Memory and Agentic AI work, explore how Neo4j enables AI agents to store and retrieve contextual information, and then build your own AI agent through a focused Mini Hack.
What is the Agent Memory Build Sprint?
The Agent Memory Build Sprint is a one day, hands on learning and building experience organised by Neo4j × hackFront India. The sprint combines technical learning, live demonstrations, mentor supported building, and a Mini Hack. Participants will first understand the fundamentals of Agent Memory and Agentic AI, then work in teams to build an AI powered agent that can remember, retrieve, and use information from its interactions. The experience is designed to move from understanding the technology to building with it and finally showcasing a working solution.
The Challenge
Build an AI Agent That Remembers
Your team will build an AI powered agent that can:
- Understand and interact with a user
- Store relevant information from interactions
- Retrieve relevant information from its memory
- Use relationships and context to improve its responses
- Demonstrate how persistent memory makes the agent more useful
The use case is completely up to your team. You can build for education, careers, customer support, research, productivity, development, travel, knowledge management, or any other meaningful application of Agent Memory.
The Core Requirement
Neo4j must play a meaningful role in the memory or relationship layer of your solution.
Your solution should demonstrate more than just an AI interface. The judges will want to see how your agent uses memory, relationships, and context to make its interactions more useful.
The Agent Memory capability is mandatory, while the specific problem statement and use case are yours to choose.
What You Will Experience
01. Understand
Start with an introduction to hackFront India and Neo4j, followed by a technical session on Agent Memory and Agentic AI. Learn how AI agents store, recall, and reason over information, why memory can be represented through relationships, and where Neo4j fits into the AI stack.
02. See It in Action
Watch a live demonstration of an AI agent with persistent memory built using Neo4j. The demonstration will give you a practical reference for the kind of Agent Memory experience you will build during the Mini Hack.
03. Build
Form a team of 2 to 4 participants and enter the two hour Build Sprint. Work on your chosen use case, build your agent, integrate Neo4j into its memory or relationship layer, and turn your concept into a working prototype. Neo4j mentors will be available during the sprint to help teams move forward and unblock technical challenges.
04. Submit
At the end of the Build Sprint, teams will submit their solutions for evaluation. Judges will shortlist the teams that move forward to the final demonstrations.
05. Showcase
Shortlisted teams will demonstrate their solutions live to the judges and audience.
The focus is on showing the memory working, not just presenting an interface. Demonstrate how your agent stores, retrieves, and uses information to improve its interaction with the user.
Event Schedule
Date: Saturday, 26 September 2026
Time: 10:00 AM to 4:00 PM
Venue: DataPhi AI Office, iSprout GreyStone, Baner, Pune