AI Portfolio
AI Marketing & Sales Team
Building my own AI go-to-market team
The Problem
Solopreneur math doesn't add up
Here's the thing about running a one-person shop: the math doesn't work.
When I'm on a client project, I'm on it. I can keep posting on social for two, maybe three weeks. After that, sales stuff just stops happening. There's no time. The client always wins.
It all feels fine until the project ends. Then I look up and my channels have gone quiet, the pipeline is empty, and I'm starting from zero. My projects tend to be long, so by the time one wraps up, the silence has been piling up for months.
So the problem isn't "can AI help me write posts." It's that delivery and pipeline can't both run on one calendar, and I kept falling off the same cliff every time a project ended.
The Goal
AI runs the go-to-market, I approve and decide
What I actually want is an AI team that does most of the work.
Marketing, social, content, outbound, follow-up, qualification. All of it running while I'm deep in a client project. Then when a prospect shows up on my calendar, everything before the meeting is already done.
My job shrinks to two things: approving what goes out, and deciding where we're pointing. Taste, tone, which deals to chase. Everything else runs without me in the middle.
That's where I'm trying to get. The rest of this page is an honest look at how far I actually am.
The Team
Agents, not tools
I set this up the way I'd set up a real team. Actual roles, not a pile of tools. Who owns what, who hands off to whom, who gets the final call.
The diagram is a snapshot. Parts of it are already a step behind the actual build, but the shape holds up. I sit at the top as the only human on the chart. Below me is a context layer, five live files that hold the strategy, ICP, content patterns, pillar plan, and the latest performance data. Every agent reads from those files before it does anything. That's how the system stays consistent without me briefing it every time.
From there it splits into two pipelines. **Content production** drafts LinkedIn posts and blog articles. Content Refinery does the writing, Humanizer cleans out the AI patterns automatically, Imagen 4 generates the cover images. **Prospecting** finds people who match the ICP, scores them, and runs them into Waalaxy sequences. **Weekly Pulse** closes the loop. Every Monday it pulls LinkedIn and GA4 data, writes it back into the context layer, and the next batch of content knows what worked.
Each agent has a clear job, its own context, and something measurable to hit. The handoffs are spelled out so nothing quietly falls through the cracks.
> **Note on what's not here:** strategy and positioning sit one level up. That's part of how I run client work and my own product thinking. It feeds into the context layer above, but it's not part of this GTM team. This page is about the marketing and sales machine.
Where I Am Now
What's live, what's next
Honest status. No point pretending on my own site.
Live and running
- **Weekly Content Batch.** Every Thursday 9am, the content agents draft 3 LinkedIn posts and 1 blog article with cover image. I review and approve.
- **Weekly Pulse.** Every Monday 9am, LinkedIn and GA4 data flows into the context layer so the agents learn what's working.
- **Five context files in production:** pulse data, proven content patterns, ICP scoring criteria, pillar-page strategy, brand voice and content strategy.
- **Blog infrastructure:** Next.js + MDX, 5 pillar hubs (3 live), automated related-articles, BioCard, MailerLite newsletter signup, content dashboard.
- **Prospecting pipeline end-to-end:** LinkedIn Search Playbook > Waalaxy import > CSV export > ICP Scorer (tiers HOT/WARM/NURTURE) > Outreach Sequences (7 title groups x 4 steps with A/B variants) > Waalaxy campaigns.
- **14 blog articles published**, 4 pillar pages drafted.
Still building
- **Prospect research.** Deep-dive enrichment on HOT prospects before outreach goes out.
- **Outreach personalization skill.** Auto-personalized messages plus article recommendations per title group.
- **Three outreach articles** still needed before the matching sequences can fully ship: business development, operations & process, and program management angles.
Still messy or open
- **Taste and tone.** How much can I hand off before it stops sounding like me.
- **Guardrails.** When should an agent ask me first versus just do the thing.
- **The diagram itself.** The real build has already moved past parts of it.
I'll keep updating this page as the team takes over more. If you want to see what this looks like in motion, or talk about building something similar for your own shop, the contact link is at the bottom.
Further reading
Growth, Growth, Growth — But From Where, Really?
Before you push growth harder, look at how buying actually works from the customer's point of view. Growth doesn't fix friction — it scales it.
The Bottleneck Isn't Building. It's Deciding.
AI made execution fast. But approval still runs on slide decks and steering groups. The bottleneck in product development isn't technical — it's organizational.
Customer Experience Has to Be Led, Not Just Talked About
CX is cross-functional by nature, but almost nobody has the authority to lead it that way. That's why the experience customers get rarely matches the ambition.
Get the next article in your inbox
Short, practical notes on strategy execution, operating models, and the structures that decide whether a transformation actually changes how teams work.