I build high-performance systems bridging local AI inference, developer tooling, and interactive simulation architectures.
- ⚡ Local AI & Edge Inference: Architecting low-latency model inference, MLX C++ engine ports, ROCm/GPU hardware acceleration, and custom quantization workflows.
- 🛠️ Developer Productivity & IDEs: Creator of SimpleSignal, bringing local models (Lemonade, Ollama, LM Studio) and cloud APIs (Qwen, DeepSeek) natively into VS Code & AI code editors.
- 🎮 Game Systems & Simulations: Engineering modular game engines, robust real-time networking, procedural systems, and data pipelines in Luau, TypeScript, and C++.
- 🧠 Autonomous Agent Pipelines: Multi-model consensus routing, tool calling integrations, and local RAG workspaces.
|
Universal AI Model Provider for VS Code. Streams Lemonade, Ollama, LM Studio, Qwen, and DeepSeek into native Chat, Inline Edit ( |
High-performance MLX C++ inference engine ported to AMD ROCm (Windows). Powers native LemonSeed, Gemma 4, and high-throughput OpenAI-compatible embedding servers. |
|
Automated quantization pipelines for Gemma4 and Snowfox architectures into ultra-compact MLX 4-bit, 6-bit, and 8-bit formats for lightning-fast edge inference. |
Large-scale modular game architectures, custom state machines, entity component systems, inventory/rebirth engines, and real-time networking sync. |
_____________________________________________________________
/ \
| "Pushing compute to the edge, one optimization at a time." |
\_____________________________________________________________/
\ ^__^
\ (oo)\_______
(__)\ )\/\
||----w |
|| ||
⭐️ Feel free to star my repositories if you find them helpful!