Skip to main content

Content Operations

AI Content Operations System

End-to-end content operations across 3 websites with distinct brand voices. From keyword research to published, SEO-optimized article in under an hour, with automated social distribution across LinkedIn, X, and Facebook.

Jonathan Lasley

Jonathan Lasley

Fractional AI Director

Claude Code Skillsn8n OrchestrationBrand Voice TrainingSocial Distribution

At a Glance

< 1 hr

Research to Published

Fully optimized article from keyword research to live post

3 sites

Distinct Brand Voices

Each site has a trained voice profile for consistent tone

2x / week

Publishing Cadence

Sustained output across all 3 sites as a solo operator


The Challenge

One Person, Three Brands, Zero Content Team

Three websites, three audiences, three brand voices, one person. Manual content production took hours per article: research, outlining, drafting, editing, SEO optimization, image creation, social promotion. At that pace, maintaining any publishing cadence across three sites was unrealistic.

The quality problem was worse than the speed problem. Consistency across brand voices required deliberate attention. A single writing session could start strong and drift. And the tasks surrounding each article: keyword research, meta descriptions, structured data, social card generation, social media posts. Those consumed as much time as the writing itself.


The Approach

Two Systems, One Process

Decision-makers: the business results are in the next section. This section covers the technical architecture your team will want to evaluate.

Not one system but two, each optimized for different requirements.

jonathanlasley.ai: Claude Code Skills (30–40 min per article)

This site uses a skill-based workflow built directly into the development environment. A /blog-research skill handles competitive keyword analysis, source gathering, and content brief generation. A /blog-write skill produces the full article: SEO-optimized MDX with frontmatter, internal cross-links, structured data, and auto-generated social cards. A /social-post skill generates LinkedIn, X, and Facebook posts at three different angles per article.

Add in n8n orchestration to finish the automated pipeline, and the entire workflow runs in 30–40 minutes. The content plan is researched monthly and stays static until the next update. Each individual article executes against that plan.

upskalr.com and thetrappedtrader.com: n8n Orchestration Pipeline

These sites use an automated n8n pipeline for content production. The workflow handles topic research, sentiment analysis, AI drafting with trained voice profiles, image generation, and CMS publishing. n8n manages the orchestration; AI models handle the creative work.

Shared Layer: Brand Voice Training

Each site has a trained voice profile regardless of which system produces the content. jonathanlasley.ai writes in my voice as an experienced AI advisor: direct, specific, numbers-driven. The other sites have their own voice characteristics matched to their audiences. The AI doesn’t produce generic content and then get manually edited for voice. It writes in the target voice from the first draft.

Social Distribution

Every article automatically generates social media content: LinkedIn, X, and Facebook posts at three angles per article. Nine social posts per article, each tailored to the platform’s format and audience expectations, and each to address specific needs of the audience.

Here’s how these pieces connect. Data sources feed into two orchestration engines, each routing through shared AI processing and quality gates before publishing to the target site.

Dual-system architecture showing Claude Code Skills for jonathanlasley.ai and n8n orchestration for upskalr.com and thetrappedtrader.com, with shared brand voice training and social distribution layers

The Results

Research to Published in Under an Hour

Dozens of articles produced across the three sites. Publishing cadence maintained at 2 per week for jonathanlasley.ai. Articles are ranking for competitive AI consulting keywords, with some reaching target positions within days of launch.

Social distribution at scale

The social distribution system has generated hundreds of platform-specific posts. Each post is written for the platform, not just the article title reposted with a link.

Brand voice consistency is measurable

Content produced at the end of a session matches the voice profile as closely as content produced at the beginning. The trained voice profiles eliminate the drift that happens in manual writing sessions.

Time comparison showing manual content production versus automated pipeline with Claude Code Skills and n8n orchestration

Why This Matters

Content Operations, Not Content Creation

Most companies produce content on an ad-hoc basis: someone writes when they have time, publishes when they remember, promotes when they think of it. Content operations replaces that with a system.

Content operations means a repeatable system with defined inputs, consistent quality, and automated distribution. A mid-market company that publishes one well-optimized article per week will outrank competitors producing occasional blog posts with no SEO strategy.

The dual-system architecture is intentional. Some content needs deep integration with the development environment (like generating structured data and social cards). Other content benefits from broader automation (like multi-site publishing). Using the right tool for each job produces better results than forcing everything through one pipeline. The brand voice profiles and prompt architectures were developed using PromptAssay, and the same prompt chain patterns that power the AI Win Strategy System drive the research and drafting skills here.

I build these systems as part of AI Strategy Assessments, and I write about content operations and AI implementation on the blog.


Key Takeaways

What Makes This Work

Brand voice training is the difference between AI content and good AI content. Generic AI output is obvious. Trained voice profiles produce content that reads like a specific person wrote it.

Dual-system architecture matches the tool to the job. Claude Code Skills for deep integration. n8n for broader automation. Trying to force both through one system produces mediocre results at both.

Social distribution should be automated, not an afterthought. Nine platform-specific posts per article, generated alongside the content, not manually written days later when the article is already stale.

Content operations scales; content creation doesn’t. A system that produces one article per week can produce two per week with the same process. Manual writing doesn’t scale that way.


Your Content Taking Days Instead of Hours?

I build content operations systems that produce SEO-optimized articles with consistent brand voice and automated social distribution. The same dual-system architecture works for any company publishing across multiple channels.

Take the Free Assessment