Content operations, AI enablement, and technical marketing for organizations that need more than a prompt and a prayer.
I build practical systems for turning expertise into credible, reusable content: governed AI workflows, quality gates, evidence checks, messaging operations, and the operating models around them.
LinkedIn · Newsletter ·
Fear, loathing, and content pipelines
A talk about the gap between “we have AI” and “this actually works in production.”
I’ll be presenting at All Day AI on October 22, 2026. The supporting repository, fear-loathing-and-content-pipelines, is where I’m collecting the practical examples, working notes, and artifacts behind the session.
The subject is not how to generate more content. It is how to create a content system that can retain judgment, expose uncertainty, enforce standards, and survive contact with real editorial work.
These are public working samples, not speculative product pages. Each demonstrates a distinct part of an AI-enabled content operating system.
| Sample | What it demonstrates |
|---|---|
| Cleartext | A voice-disciplined B2B writing environment with configurable rules packs and quality gates. |
| E-E-A-T Content Pipeline | A five-stage drafting and credibility workflow that treats evidence and specificity as hard editorial requirements. |
| Campaign Message Engine | A system for adapting one core campaign argument across six channels without flattening the voice or losing the point. |
| Content Evidence Auditor | An editorial agent that audits claims, stale statistics, weak sources, and overpromising before publication. |
| ContentOps | Open frameworks, playbooks, templates, and checklists for building a functional content operation. |
- AI content operations: Designing the intake, handoffs, governance, QA, and human review that make AI useful beyond an isolated drafting session.
- Technical and product marketing: Translating complicated technology into content and messaging that technical buyers and executives can both trust.
- Knowledge and collaboration systems: Fixing the structures, standards, and workflows behind documentation, wikis, and institutional knowledge.
- Editorial credibility: Building checks for evidence quality, specificity, claim support, and voice so “AI-assisted” does not become “smoothly unconvincing.”
Most AI content problems start before the model produces a sentence.
The request is vague. The audience is undefined. The source material is weak. Product claims drift. SMEs are handed a nearly finished draft and asked to bless it. Then everyone wonders why the output sounds generic or cannot survive scrutiny.
A better model is:
intake → evidence → brief → draft → quality gate → review → approved source → reuse
The point is not to automate editorial judgment out of existence. It is to give judgment a clearer place to operate, with fewer preventable failures reaching the end of the process.
- Content Ops Diagnostic Project
- Citation Desk
- Source Desk Agent
- DevSecOps Playbook
- CTRL+ALT Collaboration Concept
The public repositories are designed to be inspected, adapted, and challenged. Some are full frameworks; others are deliberately narrow examples of one control point in a larger system.
If you are evaluating content operations, AI enablement, technical marketing, or knowledge systems work, start with the pipeline samples above. They show how I think about the work when the real requirement is not “make content faster,” but “make the operation more reliable.”