I build ambitious local-first creative tools, strategy simulations, and game-development automation — mostly solo, with AI as a force multiplier.
I like taking ideas that normally need a team — creative-production software, persistent strategy systems, automated game testing, local AI workflows — and seeing how far one developer can push them with modern AI-assisted development.
I care less about generating lots of code and more about making the result testable, reproducible, recoverable, and actually useful.
My current private Unreal research is focused on reusable strategy-game architecture and development tooling: persistent campaign/tactical state hand-offs, deterministic simulation, automated-player missions, real input testing, visual QA/oracles, save/reload proof, and evidence-gated AI-assisted development.
The long-term goal is to make future games cheaper and faster to build by reusing the hard infrastructure instead of starting from zero each time.
Important: sponsorship supports my original development work, tooling, research, documentation, and rights-cleared projects. I do not sell access to third-party IP, copyrighted game assets, or private proprietary content.
A free public beta for local-first script-to-audio production: write or import scripts, prepare voices, generate and review takes, add emotion/SFX/ambience/music, arrange a timeline, and export a finished mix.
An experimental persistent 4X campaign and combat simulation framework with evolving factions, strategic AI, campaign economies, tactical combat, reporting, telemetry, and optional Rust/GPU acceleration.
- AI-assisted, not AI-unchecked — generated work still has to survive tests, review, and real use.
- Evidence over claims — if I say something works, I try to back it with a reproducible run or automated check.
- Reusable systems — solve the hard problem once, then carry that infrastructure into the next project.
- Local-first where practical — creative data should stay on your own machine unless you choose otherwise.
- Build in public when I can — failed approaches and rough edges are part of the process, not something to hide.
If you enjoy following the projects or want to help me keep pushing the experiments further, GitHub Sponsors is the easiest way to support me.
Sponsorship helps with development time, testing hardware, storage/build infrastructure, release work, documentation, and turning useful internal tooling into reusable public projects where licensing allows it.
There is no requirement to sponsor to use my public projects. If money is tight, a star, useful bug report, demo share, or telling somebody about the work genuinely helps too.
More detail: SUPPORT.md



