I engineer high-performance intersections between Agentic Multi-Agent Frameworks and Low-Level Game Architecture. Currently architecting production-grade AI intelligence networks and pushing the limits of real-time C++ systems.

• AI Engineer at Sofa — Architecting complex orchestration systems, production multi-agent workflows, and highly optimized local LLM deployment strategies.
• Game Developer at Nerd Magic Games — Building modular C++ core systems, advanced gameplay plugins, and robust architecture inside Unreal Engine 5.

• Orchestration & Workflow Pipelines: Specialized in constructing stateful multi-agent systems using n8n, LangChain, and OpenRouter to handle complex, non-linear reasoning chains.
• Local LLM Optimization: Deep-dive implementation of Ollama and llama.cpp environments utilizing heavily quantized model formats (GGUF) and custom VRAM memory management.
• Infrastructure Scaling: Designed and implemented a custom distributed CUDA RPC cluster topology to offload and scale high-throughput model inference across local hardware nodes.

• Plugin Architecture: Author of modular UE5 systems, building clean C++ APIs that expose optimized underlying systems to Blueprint interfaces.
• Gameplay Mechanics: Designing extensible inventory pipelines, custom character rigging automation, and low-level engine event routing.

• 🔒 Private AI Project — A sophisticated Multi-Agent Orchestrator designed for intelligent service management, routing autonomous sub-agents to process and verify real-world data structures securely. (Internal — details under NDA)
• 🔮 Nerd Magic: Technomancy — A specialized Unreal Engine 5 C++ engine plugin designed to handle advanced gameplay capabilities and modular architectural framework design for studio production.

⚙️ Languages & Runtimes

🧠 Artificial Intelligence

🎮 Engine & Environment

When compiling heavy C++ codebases or spinning up cluster-wide LLM instances, this local hardware drives the development process:

• Processor: Intel Core i9-14900KF (24 Cores / 32 Threads)
• Memory: 112GB High-Speed RAM (Optimized for large context windows & concurrent build pipelines)
• Graphics: NVIDIA GeForce RTX 4080 Super (Local VRAM management & CUDA layer execution)
• Mobile Unit: Lenovo Legion Y530 (WSL2/Ubuntu environment for target testing and on-site deployment)

📫 How to reach me: Feel free to explore my repositories or open an issue/PR if you're interested in multi-agent networks, CUDA configuration tricks, or Unreal Engine plugin mechanics.