Cobblecat · Complex Systems Design and Implementation · Python · Security · Analytics
I've led 40+ warehouse management system implementations, taking customer deployments through configuration, data migration, testing, training, and go-live. My work has involved Python and SQL, AS/400 systems, account and role configuration, and troubleshooting across warehouse software and automation. I'm a certified crime analyst, a licensed teacher, and educated in Computer Security and Information Assurance. When I'm not working... I'm working on projects.
That experience shapes my independent development and security work: understanding system behavior, protecting data integrity, and defining what users and tools should be allowed to do.
Operational Variance Investigation Toolkit · Python · SQLite · Public
A local miniature WMS and synthetic warehouse simulator with an interactive console and an operational analysis workflow. It models inventory, picks, replenishment, adjustments, and audit history, then supports independent inventory reconstruction and statistical investigation.
The architecture separates the WMS, scenario drivers, and analysis so an investigation works from operational evidence without access to the hidden answer.
Public, actively developed. Uses synthetic data; scoped as a simulation and investigation project.
Agentic gameplay aide · Python · GPT-5.6 Terra · Experimental · Private
An agentic overlay where model and tool orchestration work with persistent campaign memory to connect immediate observations with the broader playthrough. My first venture into agentic programming, it became an intensive hands on education in complex systems architecture, state management, and recovery from critical failures.
The interesting work is in the details: noticing something useful on screen and carrying the right context forward. Security design concerns include agent authority and governance, tool permissions, private session data, and safe failure recovery. Dig around in the logs and documents and discover the critical failure that almost bankrupted me.
I will bring this project public after a few necessary updates.
Access-control simulator · Python · Private
An extremely robust terminal simulator for exploring identities, groups, elevation eligibility, and Windows-style permissions. It models ACLs and inheritance to make access decisions easier to inspect and reason about.
Agentic AI system · GPT-6.1 Sol · Knowledge engineering · Research methodology · Private
Gwent Analyst is an agentic research and analysis program for GWENT: The Witcher Card Game. It combines an evolving knowledge base, a normalized rules kernel, and a proprietary structured analytical system (Axiom Engine) to support card interaction research, deck analysis, probability modeling, and detailed strategy reports. The project began as a logical knowledgebase governed by a plain-text runtime, building on my 2026 publication, Gwent: The Quantitative Strategy Manual. It has since developed into a program that carries that research methodology into software, enabling an agent to gather evidence, test strategic hypotheses, and build on validated findings, which it demonstrates though its interactions as an LLM assistant.
Science-fiction signal-operations game · Standalone · Private
A compact, standalone game about recovering information under pressure. Its first playable mission takes place on a single screen with one fictional target: Vesper Relay. Recover a Lumen shard, clear the echoes your tools leave behind, and disconnect before the link window closes or suspicion reaches its limit.
The gameplay centers on balancing progress, cleanup, and a timely exit within a tightly scoped science-fiction mission.
- Implementation ownership: coordinating small teams and taking deployments from setup through customer handoff.
- Data work: Python migration and validation workflows, SQL bulk changes, and investigating inconsistent records.
- Troubleshooting: following logs and system behavior across WMS, warehouse-control, and PLC boundaries.
- Project security work: defining threat-model scope, reviewing access boundaries, evaluating vulnerability evidence, and planning remediation with regression and usability checks.
- Current focus: systems development, agentic software development, AI governance and authority, cybersecurity, and failure recovery.