Building an AI-Powered Content Workflow — From Scratch, as a Team of One
How I designed two specialized AI agents, a buyer-journey-mapped editorial calendar, and a blog-to-pipeline system — built to scale as a one-person content operation.
How do you build a director-level content operation when you’re the only one in the room?
After 15 years building and leading content teams, I wanted to answer a question that matters to every lean marketing org: what does an enterprise-quality content workflow look like when you remove the team — but keep the standards?
Most solopreneurs and small teams default to doing everything manually or using AI generically — asking it to “write a blog post” and hoping for the best. That produces generic content at best. I wanted something different: specialized agents trained with my expertise, my voice, my standards, and my sources — operating like a focused two-person editorial team.
The result is a blog called Baking with a Marketer — an intentionally adjacent brand that pairs internet-sourced baking recipes with real marketing strategy, built to establish thought leadership in content marketing while demonstrating exactly the kind of creative brand thinking I bring to B2B clients.
Two AI agents. Two specialized roles. One editorial operation.
Rather than using AI as a generic assistant, I built two purpose-trained agents — each with a defined role, a knowledge base drawn from my professional experience, and a specific set of workflows they own. Naming them was deliberate: it signals that AI functions like a team member, not a tool.
The knowledge base behind the agents.
Both agents were trained on a structured knowledge base I built from scratch — drawing on 15 years of content marketing practice, professional frameworks, and trusted external sources. This is what separates specialized agents from generic AI output.
Reading literacy principles
Readers scan, not read. Attention is earned in the first sentence. Content is built for how people actually consume it.
Persuasion frameworks
Scarcity, social proof, authority, commitment, likeability — applied to content structure and CTAs.
Writing tips & style guide
Audience: marketing professionals. Viewpoint: trusted advisor. Language: clear, jargon-free. Structure: AI-search optimized.
AI search / GEO optimization
EEAT principles, schema markup, structured content (modular, reusable, indexed), conversational query targeting.
Intent-based topic framework
Every post has a topic, hook, differentiator, purpose, CTA, and source list — mapped to buyer intent stage before writing begins.
Trusted source library
Curated references including Content Marketing Institute and industry publications — so agents cite credibly, not generically.
From market signal to published post — every step, every owner.
The workflow is designed so every step has a clear owner — me, Edwards, or Ortiz — and nothing moves to the next stage until the current one is complete. This is the same discipline I applied to enterprise demand gen operations, applied to a one-person blog.
Market research & ICP signal identification
Using AI to research high-growth markets, identify buyer pain points, and surface demand signals. Topics must serve a real question the ICP is actively asking — not just fill a calendar slot.
Topic planning & source vetting
Edwards maps the approved topic to the editorial calendar, identifies the correct funnel stage and buyer intent type, plans the content structure, and vets sources for credibility and recency.
First draft — written in my voice
I write the post using the topic framework: topic, hook, differentiator, purpose. AI accelerates research but the perspective, POV, and voice are mine. This is non-negotiable — generic AI content doesn’t build thought leadership.
Copyediting pass
Edwards reviews against the full knowledge base: reading literacy principles, persuasion structure, style guide compliance, CTA effectiveness. Suggests specific edits — doesn’t rewrite, refines.
SEO & AI search optimization
Ortiz scans for keyword opportunities, adjusts structure for AI search discoverability, produces schema markup, generates Yoast meta fields, and suggests an optimized title. GEO strategy applied at this stage.
Publish & distribute
Post publishes to WordPress blog. Edwards repurposes into a LinkedIn native article with trackable UTM links back to the blog. Distribution tracked via GA4 with campaign URL builder for channel attribution.
Performance reporting
Ortiz connects to GA4 to analyze traffic, engagement, and conversion signals. Data feeds back into the editorial calendar to inform future topic prioritization and content investment decisions.
A buyer journey built specifically for the blog — not borrowed from a template.
The editorial calendar isn’t organized by month or topic category — it’s organized by reader stage. Every piece of content has a defined role in moving a reader from first discovery to brand advocate.
Reader challenge: Content marketing is rapidly changing. Where do I start?
Content typeBlog posts, LinkedIn articles, short-form social — primarily educational and thought leadership
ChannelsLinkedIn, Instagram (Baking with a Marketer), Google search
Goal: Build brand with niche, establish expertise through consistent analytics-informed content
Content typeSubscription-based content, deeper how-to guides, podcast (planned expansion)
ChannelsLinkedIn, Substack, search, Instagram
Goal: Turn readers into brand advocates — sharing content, collaborating, referring
Content typeCommunity content, collaborations, industry event appearances
ChannelsEvents, webinars, website, word of mouth
The workflow in action — real posts, real topics, real intent mapping.
Each post goes through the full Edwards → Ortiz pipeline before publish. These examples show how intent-based topic research translates to content that answers real ICP questions.
Why Data Sources Matter in AI Search
Addresses the gap in how marketers understand LLM data sourcing — a topic with almost no unbiased coverage. Targets informational intent for marketing professionals navigating AI search strategy.
Informational intent6 Tips to Make Your Content Stand Out in AI Search
Practical, actionable guidance mapped to the consideration stage — reader knows AI search matters, now needs to know what to do about it.
Commercial intent5 KPI Signals That Predict Pipeline Growth: A Revenue-Driven Framework
Speaks directly to revenue-accountability anxiety among content marketers — a high-intent topic for senior marketers being asked to prove content ROI.
Transactional intentThe style guide Edwards enforces on every post.
Marketing professionals and readers interested in marketing topics — from practitioners to senior leaders
Writing from the perspective of a trusted advisor with 15 years in the marketing field — not a lecturer, a peer
Good business language — clear, direct, no unnecessary jargon. Easy to understand and remember
AI-search and SEO optimized — scannable, modular, with clear H1/H2/H3 hierarchy and schema-ready formatting
Johanna has an excellent temperament — calm, thoughtful, and steady under pressure — and navigates complex stakeholder and personality dynamics with ease. She brings the same discipline to systems as she does to strategy: everything she builds is designed to be understood, repeated, and scaled by the people around her.