Writing systems

Writing and systems that make judgment more inspectable.

The public-thinking layer matters because management judgment improves when claims, assumptions, sources, and trade-offs are easier to inspect.

Back to the route map
01

Notes

Use notes to preserve decisions, assumptions, evidence, and unresolved questions across longer work.

02

Writing

Turn internal lessons into public-safe writing about AI quality, accountability, and engineering leadership.

03

Knowledge

Keep reusable context in structured artifacts instead of depending on repeated oral history.

04

AI Thinking

Use AI as a review partner for assumptions, counterarguments, wording, and verification plans.

05

Small Tools

Build lightweight workflows that reduce repeated coordination, synthesis, and follow-up work.

06

Evidence

Separate verified facts, inferences, and claims that still need public-safe confirmation.

Writing and notes

I use writing to catch weak assumptions early and turn LinkedIn posts, notes, and private context into clearer public-safe arguments.

  • Local versions of selected LinkedIn posts about AI quality, leadership, incentives, and communication.
  • Knowledge systems that preserve evidence instead of flattening it.
  • Drafts that keep one central claim instead of turning into generic leadership checklists.

Knowledge systems

I use notes, trackers, and small automations to make decisions, assumptions, and follow-up visible across meetings and handoffs.

  • Source-grounded notes for leadership, AI, quality, and delivery work.
  • Lightweight tools that reduce repeated context loading.
  • Explicit separation between verified evidence, inference, and public claims.

Public themes

The strongest public writing themes mirror the work evidence: AI proof, accountability language, and using multiple perspectives to improve judgment.

  • One central claim beats a generic leadership checklist.
  • A concrete workplace story is stronger than abstract thought leadership.
  • Public writing should preserve nuance without leaking internal context.