Building in public: how AI agents run our content factory

Description and use

Arxitek uses 19 AI agents for routine content tasks while humans oversee strategy, review, and final decisions, improving speed and quality.

Problem

Automating content production while maintaining quality, tone consistency, and factual accuracy.

Outcome

The system produces 77 pieces of content per week across multiple languages, with AI handling routine tasks and humans ensuring quality, tone, and factual accuracy.

Workflow

• 1. Review analytics for stalled pages, missed queries, and competitor coverage.
• 2. Researcher agent gathers topics and sources.
• 3. Writer agent drafts articles based on structured inputs.
• 4. Page builder agent generates SEO pages from templates.
• 5. Analyst agent monitors data and flags issues.
• 6. Sales assistant agent handles routine conversations.
• 7. Human editor reviews drafts, adjusts prompts and rules, and publishes content.
• 8. Post-publication analysis informs prompt and rule adjustments.

Evidence

In our content factory, AI agents draft articles, build SEO pages, support sales conversations and report on analytics, while people set strategy, review the output and make the final call.

Source

They pass structured handoffs, not vibes. When something goes wrong, we know which link broke.

Source

Agents drift in tone. They repeat themselves across articles. They sound confident about facts they can't back up.

Source

Automation isn't set-and-forget. It's maintenance, like any other system.

Source