Skip to content
View kidshadow79's full-sized avatar
  • France

Block or report kidshadow79

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
kidshadow79/README.md

🇫🇷 Version française

Yohan BROCARD

I am Yohan Brocard, a passionate autodidact. I approach software development not as a technical showcase, but as a space for personal research and questioning, particularly around artificial intelligence systems. My approach is entirely hands-on and experimental: exploring concepts, testing ideas in tandem with AI, observing and guiding the behavior of AI models, and continuously iterating to bring intriguing ideas to life.

Since May 2025, I've been building OGMA — a conversational AI assistant with persistent memory — with the help of coding AIs, learning as I go. What I didn't know how to do yesterday, I understand today. What I don't understand yet, I'll explore tomorrow.


What interests me

I don't come from the tech world. What led me to code were questions:

  • What would allow an AI to have a stable identity over time?
  • How do you build true memory — not just an extended context, but structured recollections that persist and evolve?
  • Can you design an AI that knows how to say no, that adapts its alignment when the user's genuine interest calls for it?
  • What does an ethical and authentic relationship between a human and a machine look like?

I don't have the answers. OGMA is the proving ground where I look for them.


My main project

🧠 OGMA — AI assistant with persistent memory and stable identity

Inspired by Ogmios, the Gaulish god of eloquence and communication.

OGMA explores what happens when you treat an AI as a developing entity rather than a tool. Dual architecture (conversational AI + analytical Archivist), hybrid SQLite + FAISS memory, boolean ego flag system, and memory consolidation during inactivity.

The code is a work in progress — it's monolithic and bears the marks of a learning journey. What matters is the behavior this architecture produces and the questions it raises.


What I'm looking for

I work alone, without a developer network. If you work on related topics — memory in LLMs, identity, AI ethics, or conversational systems — your perspective is truly valuable to me.

No formal collaboration needed. A remark, a critique, or a simple exchange is enough.


Contact


May 2025 — today. Permanently under construction.

Pinned Loading

  1. Ogma Ogma Public

    Assistant IA à mémoire persistante et croissance organique

    Python 5