I Built an AI Team That Runs My Q2 Content Calendar
- Jonell Zimmerman
- Mar 30
- 3 min read
I need to tell you something that's going to sound either impressive or unhinged, depending on where you are with AI right now.
I built an AI agent team. Not one assistant. A team. Multiple agents, each with a specific job, running my content calendar every week with almost no manual input from me.
And I need to tell you what actually happened. Because it wasn't the clean, triumphant story I thought it would be.
How this started.
About two months ago, I was drowning in content production. I run a consulting practice. I write a newsletter. I show up on LinkedIn five days a week. I have a podcast in development and a course I'm building. All of that, solo.
My Monday mornings were four hours minimum just planning what to say that week. Researching. Drafting. Editing. Scheduling. By the time I finished producing content, I had no energy left for the work the content was supposed to attract.
So I did what any reasonable person would do. I decided to build an AI system to handle as much of it as possible.
Not one chatbot. A whole ecosystem. An agent that does weekly research. An agent that writes strategy briefs. Agents that draft LinkedIn posts, produce newsletter outlines, handle analytics. Each one with specific instructions, specific memory files, and a specific job.
I named them. Lucy handles marketing. Connie runs operations. Bella builds products. I gave them personalities because... honestly, because it helped me think about what each role needed to do. More on that in a second.
What broke.
Here's the part you don't see in the AI highlight reels.
The first version was terrible. Not because the AI was bad. Because my instructions were bad. I gave vague directions and expected precision. I said things like "write a good LinkedIn post" and was frustrated when the output sounded like a press release written by a robot who'd read too many marketing blogs.
Sound familiar? It should. It's the exact same problem I see in every organization I consult with. Leaders give vague direction and then wonder why their teams aren't aligned.
The AI agents needed the same thing my human clients need. Clarity.
They needed to know: Who are you writing for? What does this person care about? What's the brand voice? What are the boundaries? What does "good" look like, specifically?
When I finally sat down and wrote detailed instructions for each agent, with examples, with constraints, with a clear definition of what success looks like, the output quality jumped. Not to perfect. But to a starting point I could actually work with in fifteen minutes instead of building from scratch in three hours.
What surprised me.
The thing I did not expect: the bottleneck was never the AI. It was always me.
Every time the system stalled, I could trace it back to one of three things. Missing context I forgot to provide. A decision I hadn't made yet about who I was writing for. Or unclear priorities for the week.
I started calling it clarity debt. The accumulated cost of decisions I hadn't made and context I hadn't documented. My AI agents couldn't do their jobs because I hadn't done mine.
And that hit different. Because I spend my professional life helping leaders see this pattern in their organizations. Turns out it applies to a one-woman AI operation too.
It's a design problem, not a people problem. Even when the "people" are AI.
Where it is now.
My Monday content production went from four hours to about forty minutes of review and editing. The agents draft. I decide. I refine. I approve. The system handles research, first drafts, scheduling, and analytics tracking.
Is it perfect? No. I still catch things. The AI occasionally produces something that's technically correct but emotionally flat. Or it leans too hard on a concept from last week because the memory files haven't been updated. Small shifts, small fixes. Ongoing.
But here's what changed: I have my mornings back. The hours between 9am and noon used to disappear into content logistics. Now they're spent on actual client work and building new things. That math matters when you're solo.
Why this matters for you.
You don't need to build what I built. That's not the point.
The point is this: AI works the way your team works. Give it vague instructions, get vague output. Give it clear context, specific constraints, and a defined version of "done," and it performs. The same leadership skills that make your human team better make your AI tools better.
If your Q2 plan involves AI in any capacity, start with the clarity, not the tools.
-Jonelle
P.S. If you've tried giving AI a task and the output was so generic it could have been written by literally anyone on earth, reply and tell me. 90% of bad AI output is a clarity problem in disguise.
Comments