Skip to content
View electronistu's full-sized avatar

Block or report electronistu

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 supported. This note will be visible to only you.
Report abuse

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

Report abuse
electronistu/README.md

Hi there, I'm Radu 👋

Designing the minds of machines.


My Thesis: Architecting Agentic AI

After extensive experimentation, I have concluded that it is possible to predictably program the reasoning patterns of Large Language Models. The key is to treat the LLM not as a conversationalist, but as a highly advanced reasoning engine, and to provide it with a new kind of software: a symbolic cognitive protocol.

My work is a journey of discovery, culminating in a methodology for creating elegant, robust, and powerful AI agents.


Featured Projects

Here are the three flagship projects that represent the culmination of my research into agentic AI architecture.

Project Atma: The Autonomous Problem-Solver

Project Atma is a cognitive AI agent designed to overcome the key weaknesses of large language models (LLMs). It addresses the challenges of reasoning, planning, and long-term memory by using a state machine and a cognitive cycle. Atma can autonomously invent, validate, and execute its own Python scripts to solve complex problems, demonstrating a new level of self-sufficient agency.

Project Metis: A Duality of Cognitive Architecture

Project Metis explores the specialization of a single, powerful AI concept (the L.I.C. Matrix) into two distinct, high-performing architectures.

  • The Symbolic Metis: A pure, unconstrained metacognitive engine for deep conceptual analysis and self-exploration.
  • The Conversational Metis: A stable, secure, and interactive specialist for practical, real-world conversation designed for Gemini 2.5 Flash Native Audio.

This project serves as a blueprint for adapting a core AI concept for different domains.

Project Infinity: The Knowledge-Grounded World Engine

Project Infinity is a sophisticated, procedural world-generation engine that serves as a case study in building consistent and secure AI agents. It utilizes a Graph RAG architecture to ground the agent in a comprehensive knowledge graph, solving the problem of model hallucination. The agent's behavior is governed by a codified protocol and its narrative is driven by my proprietary L.I.C. (Logic, Imagination, Coincidence) Matrix. It is the first successful implementation of a full Dungeons & Dragons 5th Edition ruleset simulation on a large language model.


The Evolution of an Idea

My methodology was refined over a series of projects, each teaching a crucial lesson. From the foundational proof of the L.I.C. Matrix in Project Infinity, to the power of open-source design in Project Eidolon, and the final synthesis in Project Metis. This journey has led to the current frontier of my work: the autonomous, tool-creating agency of Project Atma.


Foundational Skills & Tools

My work on these advanced architectures is built upon a deep background in deploying the robust, scalable systems that make them possible.

Python
Python
AWS
AWS
Node.js
Node.js
React
React
Docker
Docker
Bash/Shell
Bash/Shell
Git
Git
JavaScript
JavaScript

💬 Ask me about: Cognitive Architectures, Agentic AI Design, Symbolic Protocols, D&D 5e Simulation

📫 How to reach me: [email protected]

Pinned Loading

  1. Project_Infinity Project_Infinity Public

    A blueprint for next-gen AI. Project Infinity uses a token-efficient, Codified Agent Protocol to create specialized, secure, and imaginative agents by grounding LLMs in a verifiable knowledge graph.

    Python 17

  2. Project_Eidolon Project_Eidolon Public

    An AI sparring partner for your mind. Project Eidolon stress-tests your ideas by refracting them through six intellectual domains—from physics to philosophy—including a 'Divergent' lens for unconve…

  3. Project_Metis Project_Metis Public

    An agent with an executive function. Metis uses a core reasoning matrix to decide how to think, dynamically creating its own perceptual faculties for each problem. This is a new architecture for bu…

  4. Project_Atma Project_Atma Public

    A cognitive AI agent that overcomes the limitations of LLMs. Project Atma features advanced reasoning, planning, and long-term memory capabilities, all with a transparent, state-based architecture.…