Building N4K48: My Journey with MyZubster, AI and Neon Plaza
I am building N4K48, my digital identity and personal development path inside the MyZubster ecosystem.
N4K48 is more than a profile picture or a fictional character. It is a way to connect my technical work, experiments with artificial intelligence, open-source contributions and the visual world of Neon Plaza.
My goal is to build everything in public and clearly separate working software from concepts, future plans and visual storytelling.
What is N4K48?
N4K48 is designed as a persistent and recognizable identity.
Instead of creating a temporary guest character every time someone enters the experience, the system connects the character to an authenticated account. The server—not the browser—determines the verified identity.
The current flow is:
Identity → Authentication → Persistent Profile → Neon Plaza → AI
This foundation can later support shared spaces, interactions, observations and AI-assisted experiences.
What we have completed
The following components have been implemented and tested:
- a persistent N4K48 character profile;
- JWT-based authentication;
- a clear distinction between guests and authenticated accounts;
- authenticated access to Neon Plaza;
- server-side identity protection;
- rejection of invalid tokens instead of silently treating them as guest sessions;
- automated coverage of the authenticated flow;
- public documentation and a development roadmap.
The latest recorded verification completed:
Test Suites: 4 passed
Tests: 22 passed
Tests failed: 0
These results verify the tested local behavior. They do not mean that the complete platform is already production-ready.
Two connected repositories
My work is currently documented across two public repositories.
MyZubster development fork
The main repository contains the authenticated Metaverse flow, persistent profile, roadmap and detailed N4K48 documentation:
👉 nicolaususnicola-lgtm/myzubster
You can also read the public roadmap:
And the detailed project documentation:
Independent MyZubster MVP
The second repository explores a smaller technical MVP with:
- an API for recording observations;
- configurable JSON persistence;
- automated end-to-end tests;
- Docker support and health checks;
- local AI with Ollama;
- vector search with Qdrant;
- retrieval-augmented generation based on stored observations.
👉 nicolaususnicola-lgtm/myzubster-mvp
AI and Zorgax
Zorgax represents the AI and operational guidance layer of the project.
The purpose of AI is not to replace verification. It should help organize information, retrieve relevant observations, explain results and guide users through a workflow.
The principle is simple:
Observe → Document → Connect → Measure → Verify → Publish
Visuals and storytelling explain the journey. Code, tests, commits and evidence demonstrate what has actually been completed.
The product direction
The broader idea is to explore how AI can assist users inside MyZubster, including future workflows connected to digital identity, verified activity and value exchange.
Any payment or reward component must remain transparent and evidence-based. A contribution, issue, pull request or internal ledger entry is not automatically proof of an external payment.
This distinction is important when building an open and trustworthy product.
What comes next?
The next development priorities are:
- complete GitHub OAuth in the frontend;
- implement logout and session-expiration handling;
- test Neon Plaza with multiple simultaneous sessions;
- verify presence and online-user counting;
- complete security and privacy checks;
- deploy a reproducible staging demo.
N4K48 is still an MVP under development, but the technical foundation, documentation and public identity are now in place.
Follow the journey
You can follow my progress through my GitHub profile:
👉 github.com/nicolaususnicola-lgtm
I will continue documenting what works, what still needs validation and what I learn while building N4K48 with MyZubster and AI.
If you are interested in open source, AI, authenticated digital identities or experimental virtual environments, feedback and technical discussion are welcome.

Top comments (0)