Let me begin by saying that I am not a traditional builder with a traditional background. From the onset of this endeavor until today it has just been me, my laptop, and my ideas I learned how systems work through trial and error, and I built these platforms because after an exhaustive search I discovered a need. I am fully aware that a 54 year old fantasy novelist with no formal training creating one experimental platform, let alone three, in his kitchen, on a commercial grade Dell stretches credulity to the limits (or beyond). But I am hoping that my work speaks for itself. Although admittedly, it might speak to my insane bullheadedness and unwillingness to give up on an idea. So, if you are thinking I am delusional, I allow for that possibility. But I sure as hell hope not.
With that said
I have released three large software systems that I have been developing privately. These projects were built as a solo effort, outside institutional or commercial backing, and are now being made available, partly in the interest of transparency, preservation, and possible collaboration. But mostly because someone like me struggles to find the funding needed to bring projects of this scale to production.
All three platforms are real, open-source, deployable systems. They install via Docker, Helm, or Kubernetes, start successfully, and produce observable results.
The Platforms
ASE — Autonomous Software Engineering System
ASE is a closed-loop code creation, monitoring, and self-improving platform intended to automate and standardize parts of the software development lifecycle.
It attempts to:
• produce software artifacts from high-level tasks
• monitor the results of what it creates
• evaluate outcomes
• feed corrections back into the process
• iterate over time
ASE runs today, but the agents still require tuning, some features remain incomplete, and output quality varies depending on configuration.
VulcanAMI — Transformer / Neuro-Symbolic Hybrid AI Platform
Vulcan is an AI system built around a hybrid architecture combining transformer-based language modeling with structured reasoning and control mechanisms.
Its purpose is to address limitations of purely statistical language models by incorporating symbolic components, orchestration logic, and system-level governance.
The system deploys and operates, but reliable transformer integration remains a major engineering challenge, and significant work is still required before it could be considered robust.
FEMS — Finite Enormity Engine
Practical Multiverse Simulation Platform
FEMS is a computational platform for large-scale scenario exploration through multiverse simulation, counterfactual analysis, and causal modeling.
It is intended as a practical implementation of techniques that are often confined to research environments.
Current Status
All three systems are:
• deployable
• operational
• complex
• incomplete
What This Release Is — and Is Not
This is:
• a set of deployable foundations
• a snapshot of ongoing independent work
• an invitation for exploration, critique, and contribution
• a record of what has been built so far
This is not:
• a finished product suite
• a turnkey solution for any domain
• a claim of breakthrough performance
• a guarantee of support, polish, or roadmap execution
If you find parts that are useful, interesting, or worth improving, you are free to build on them under the terms of the license.
I know the story sounds unlikely. That is why I am not asking anyone to accept it on faith.
You can judge for yourself
— Brian D. Anderson
ASE: https://github.com/musicmonk42/The_Code_Factory_Working_V2.git
VulcanAMI: https://github.com/musicmonk42/VulcanAMI_LLM.git
FEMS: https://github.com/musicmonk42/FEMS.git