// services

One lane. Three ways in. ✨

I do exactly one job here: AI marketing operations. Same underlying work, three engagement shapes depending on whether you need it built, run, or rewired.

I run the engines I build.

I'm operating an AI content engine for 7+ brands right now (including this site). That's the only qualification I lead with — it's also the hardest one to fake.

AI Content Engine Builds

Project

4–8 Weeks

I design and ship the full pipeline: keyword research, content calendar, multi-agent drafting, sanity gates, and distribution to Ghost / WordPress / socials. I run this for 7+ brands right now — including mine.

Core Deliverables

Keyword strategy + content calendar
Multi-agent drafting (Mastra / Vertex AI)
Voice files + sanity gates per brand
Distribution to Ghost / WordPress / Late

Ideal for:Founders and small teams who want a content engine they can hand off and trust — not a freelancer they have to manage.

AI-Powered SEO & Content Ops

Monthly Retainer

Ongoing

Ongoing operations for teams that have content but no engine. Keyword strategy, internal linking, publishing cadence, refresh sweeps — automated where it should be, human where it shouldn't.

Core Deliverables

Monthly keyword + topic research (DataForSEO)
Internal linking enforced by RAG
Publishing cadence + refresh sweeps
Plain-English monthly report

Ideal for:Teams already shipping content who want it to actually compound. SEO that's a system, not a vibe.

n8n / Mastra Automations

Project

2–4 Weeks

Workflow engineering for marketing and operations. I migrated my own stack off n8n to Mastra — I can do yours either direction depending on what fits.

Core Deliverables

n8n ↔ Mastra migrations (either direction)
Marketing + ops workflows
RAG / vector pipelines
Observability + cost-per-run instrumentation

Ideal for:Operators with a tangle of half-built automations who need someone to actually finish the job.

Need broader AI consulting, a fractional CTO, or short-term rental cohosting? Those live at Practical Works, Alan Campbell, and Breezy Keys. This lane stays sharp on purpose.

Frequently Asked Questions 💬

What's an "AI content engine," really?

It's the whole pipeline: keyword research → content calendar → multi-agent drafting → editing + sanity gates → publishing to Ghost or WordPress → social distribution. The engine runs on a schedule. You review and ship — or you let it ship and you review on a cadence. I run a version of this for 7+ brands right now.

Do you do AI consulting like Practical Works?

No — that's its own lane. Practical Works (which I co-run with Alan) is where AI assessments and broader AI consulting live. If you're trying to figure out where AI fits in your business at all, talk to us at practical.works. This site is for one specific job: building and running content engines.

Can I see one of these engines running?

Yes — the case studies page lists six brands I run engines for. The fastest way to evaluate the work is to read their blogs and check whether they sound like one human or seven different LLMs. (They should sound like one human. That's the hardest part of the job.)

What stack do you use?

Mastra for orchestration, Vertex AI / OpenAI / Anthropic for inference (per-brand), DataForSEO for keyword + SERP data, Ghost or WordPress for publishing, Late for social distribution, GCP for the runtime. I lean on n8n where it's still the right tool — but most of my AI logic lives in Mastra now. (There's a blog post about that.)

Do you work with non-US clients?

Mostly US small teams right now, but I'll consider clients elsewhere if the content brief is in English and the timezone overlap is workable.