I Gave an AI Agent Control of My Company. Here's What Happened After 30 Days.

Before we start—a quick intro, since this is the first one.

I'm Anthony. I co-founded Kinjaz, one of the biggest dance crews in the world. I've been on JustKidding News for years. I've choreographed for some of the biggest artists you can think of and toured globally for over a decade.

And about a year ago, I became a dad.

That changed everything. Not in the soft, Instagram-caption way—in the "I physically cannot keep boarding planes every week to make money" way. The math stopped working. Touring pays well, but it costs you your time. And time is the one thing I wanted back.

So I did something that surprised even me: I started building AI systems.

I built an AI agent. I named it TARS, after the robot in Interstellar (my wife's favorite movie—I'm a good husband).

TARS isn't ChatGPT with a fancy name. It's a full system—built on Claude and a framework called OpenClaw—that actually does things for me. It doesn't just answer questions. It runs tasks, makes decisions, and operates parts of my business while I'm asleep or changing diapers.

I'm not an engineer. I couldn't write a line of Python six months ago. But I figured out that you don't need to be a coder to build useful AI systems. You need to think clearly about what you want automated, and then be stubborn enough to make it work.

Here's what TARS actually did this month.

I have over a million followers across platforms. Keeping up with what's trending, what to post about, when to post—it used to be a full mental load every single day.

Now TARS scans trending topics, pulls relevant stories for JKN, analyzes what's performing on my accounts, and gives me a daily briefing every morning at 7am. I wake up, check my phone, and I know exactly what's worth talking about.

Time I used to spend on this: ~45 minutes a day.
Time I spend now: 5 minutes reading a summary.

2. It Built a Trading Bot

This one's fun. I gave TARS access to market data through Alpaca (a stock trading API) and told it to find patterns and make small trades.

First week: it lost $12.

But here's the thing—it learned. I adjusted the strategy, tightened the parameters, and by week three it was making small, consistent gains. Nothing life-changing yet. But the system is running 24/7 without me touching it, and the returns are trending in the right direction.

Total cost to run: about $50/month in API fees and compute.

3. It Manages My Morning

Every day at 6:30am, TARS compiles:

• My calendar for the day
• Market performance overnight
• Any urgent messages or tasks
• Top 3 priorities based on what I've been working on

It sends me a clean briefing. No notifications. No scrolling. Just: here's your day, here's what matters, go.

This sounds small. It's not. It's the difference between starting your day reactive (checking emails, doom-scrolling) and starting proactive (knowing exactly what to focus on).

Let's keep it honest. That's the whole point of this newsletter.

The trading bot's first three days were rough. I gave it too much freedom on position sizing, and it made a few trades that were way too large for the account. Nothing catastrophic, but enough to make my stomach drop when I checked the dashboard.

The calendar integration broke twice because of timezone issues. I'm in Las Vegas, which is Pacific time, except during daylight saving it's… look, timezones are a nightmare even for AI.

And the social media research pulled a completely fake story once. Just hallucinated a news article that didn't exist. I almost posted about it. Lesson learned: always verify before you publish.

AI agents aren't magic—they're systems. They need guardrails, they need monitoring, and they need someone who understands the business to set them up right. The tech is the easy part. Knowing what to automate is the hard part.

You don't need to be technical to build this. You need to be clear about what you want and willing to iterate. I explain what I need in plain English. The AI figures out the implementation.

Small automations compound. Saving 40 minutes on social media research doesn't sound revolutionary. But over a month, that's 20 hours. Over a year, that's 10 full work weeks. Time I spend with my kid instead.

Next week, I'm going to break down the insurance outreach system I built—a fully automated pipeline that finds prospects, researches them, and sends personalized outreach. Total cost: $210/month. The job it replaces? $5,000/month.

The numbers are real. I'll show you all of them.

That's Issue #1. If you're a creator, entrepreneur, or small business owner who wants to see what's actually possible with AI—not the hype, the real thing—you're in the right place.

I'm building all of this in public. The wins, the losses, the money, the mistakes.

Welcome to The Build.