Problems I could not leave alone.

The thesis: Enterprise AI works when routine steps stay deterministic and intelligence is used only where judgment adds value.

Scalong's enterprise AI agent layer for running work across systems, teams, and long-lived processes. It combines deterministic workflows with AI at decision points, with approvals, recovery, and auditability built in.

The thesis: When AI accelerates code production, delivery control becomes the new bottleneck.

An orchestration thesis spanning requirements, QA, UAT, delivery risk, and release evidence, designed to work across the tools teams already use.

The thesis: More autonomous coding agents need stronger stopping conditions, not merely longer prompts.

A portable workflow for Claude Code and Codex that turns an ambiguous task into a work contract, traces the end-to-end flow, and makes validation evidence explicit.

The thesis: AI is creating more small software than existing distribution systems are designed to handle.

A distribution layer where creators publish from their coding agent, then add optional hosting, secure access, shareable links, and commerce without rebuilding the product around infrastructure.

The thesis: Small workflow improvements compound when they remove friction you feel every day.

A small developer tool born from repeated friction in my own workflow, built to keep useful commands organized and close to the work.