Games and simulations for humans, robots and AI agents.
Building world models where humans, robots and AI agents learn, play, and evolve together.
Shield AI is a combination of awesome hardware, future tech and personal responsibility. And I can’t imagine anything harder and more compelling than protecting people with intelligent systems.
Managing complexity by carefully designing simple axiomatic building blocks. Then you can let your robot loose in an open world where you're encouraged to experiment.
In the two videos he has posted we see multiple weapons, using missiles to break things, a grappling hook to solve physics puzzles, and many other amazing things that have come out of this one-man wolfpack of iPhone development.
For a casual card game, Briscas looks fantastic. The game is fully animated, and includes special effects like lightning bolts and earthquakes for winning big tricks.
October 2025
Maquina is a framework for rapidly building digital twins using three simple primitives (Objects, Operations, Machines), enabling LLMs to create personalized simulations while providing structured world models for AI agents.
July 2025
What if the secret to creating vibrant virtual worlds isn't achieving perfect balance, but measuring how far we are from it? A new perspective on using game theory in virtual world design.
April 2025
Introducing Situation Design: a higher-level framework for managing the exponential complexity of dynamic interactions in massive games through organic, scalable "bubbles" of activity.
February 2025
Revolutionizing virtual worlds by starting with meaning instead of objects, and implementing distributed environmental memory that creates truly persistent, organic worlds.
December 2024
Exploring how ModernBERT's efficient architecture and strong classification capabilities can enhance NPC behavior, dynamic difficulty, and procedural content validation in games.
December 2024
From Actor Models to Behavior Trees, explore the cutting-edge AI architectures that enable virtual worlds with entities that learn, adapt, and create emergent behaviors.
AI Engineer for Games and Simulations
I'm interested in crazy ideas, body tech upgrades, life extension, curing cancer; or creating virtual worlds where none of this matters.
Unreal Engine is my engine of choice. Using physics, AI and control systems for building immersive experiences for eXtended Realities.
Over 20 years building games, simulations, and real-time interactive systems across Unreal Engine, Unity, and custom pipelines. My current focus is simulation-first AI: agentic systems, synthetic data, digital twins, robotics perception, and sim-to-real workflows for worlds where humans, robots, and AI agents can train, test, and interact.
I work across simulation, AI, and systems programming: Unreal and Unity for real-time worlds, Python and PyTorch for AI workflows, and Rust with Candle where reliability, performance, and strong abstractions matter.
Tools and systems I use to build interactive worlds where agents perceive, learn, act, and adapt.
Python • C++ • Rust • C# • CUDA • GLSL/HLSL • TypeScript • Lua
PyTorch • Candle • Hugging Face • Transformers • LLM tool use • Function calling • RAG • Embeddings • Vector search • Multimodal AI • Agent orchestration • Evaluation pipelines
Unreal Engine • Unity • NVIDIA Isaac Sim • Isaac Lab • ROS 2 • OpenUSD / Omniverse • Gazebo • MuJoCo • Reinforcement learning • Imitation learning • Domain randomization • Sim-to-real transfer • Digital twins • Synthetic data
OpenCV • GStreamer • TensorRT / ONNX Runtime • Sensor fusion • LiDAR • RGB-D • Pose estimation • Segmentation • Video analytics
OpenXR • Meta Quest • SteamVR • WebGPU • WebRTC • Multiplayer simulation • Procedural generation • Autonomous NPCs • Human-in-the-loop tools
