What I am learning by building.
Practical notes on product decisions, AI-assisted workflows, Customer Success systems, experiments, failures, fixes, and the methods that survive another iteration. Working behavior and future ideas stay clearly separated.
Build the Operating System, Then Make It Reusable
How Customer Success operating experience became a reusable CRM-building skill with onboarding, stakeholder workflows, approval-gated AI, Exa research, and least-privilege MCP integrations.
Beyond Auto-Compaction: Quillgeist Moves to Delta-State Context
Why independent convergence with dlab’s CliffCompaction pushed Quillgeist from recursive summaries toward exact anchors, bounded working state, cold history, and just-in-time rehydration.
Think-Sculpting: When Thinking Becomes Building
How AI compresses the distance between an idea and a working test, and why faster building requires stronger guardrails, verification, and iteration.
Getting the Teams Background Just Right
How a specific visual requirement became a reusable background-generation workflow and public tool.
Building the Next Step
The builder-origin story behind Clintware: start with a real operating problem, build the smallest useful system, test it, and keep improving it.