/ case studies / the operator Work with me

Case Study · The Operator Behind the Case Studies

Run your life and business with AI.

I’m Peter. I run a consulting practice, a web studio, a daily content operation, and my own investing — and AI runs the middle of all of it. Not someday. Today, most of it while I sleep. This isn’t a demo I built to sell you; it’s the place where every method I bring to a client gets tested on my own operation first. Here are the receipts.

Operator: One person Runs: 9 live sites · 1.18M pages On autopilot: 36 automations Cost to run: under $400 / month
9
Live sites I run, plus a graveyard of redirects I’ve consolidated into them
1.18M
Pages built and maintained by agents — raceteam.wiki alone is 1.12M teams
2 / day
Videos that render and publish themselves, unattended, every single day
500+
Consecutive days my agents have briefed me every morning before I’m up

A day in the operation

What happens while I sleep.

Here’s a real 24 hours. The indigo steps run themselves — on a schedule, in the cloud, whether my laptop is open or not. The green steps are the handful of places I actually show up.

11:00 PM

The operation backs itself upagent

Repositories and databases are archived off-site before the day’s work is even finished. Nothing important lives in only one place.

2:00 AM

It checks its own pulse — and healsagent

A nightly sweep pings every site, confirms the latest content is live, re-dispatches failed jobs, and repairs what drifted. It doesn’t wait for a person.

4:00 AM

Four agents read the news and argueagent

My morning council debates the day’s AI and tech news from different angles, then writes one cross-signal briefing. It’s done this for 500+ days straight.

4:00 AM

A briefing video renders and uploads itselfagent

Script, voice, visuals, render, upload — the Daily Intel Briefing is on YouTube before dawn. I never touch it.

4:45 AM

A second video publishesagent

“The Arena” has shipped a fresh short every single day since March. That’s two finished videos posted before I’ve opened my eyes.

~8:00 AM

I wake up to a finished briefingme

My first act of the day is reading, not doing. The research, the monitoring, the overnight content — already handled. I decide what matters.

Daytime

I do the work only I can dome

Client discovery, strategy, judgment calls, relationships. The agents prepped the dashboards, the docs, and the process maps. I bring the thinking.

Afternoon

Client sites update themselvesagent

An owner emails a change; an agent drafts it; I approve; it ships to their domain. That’s a whole web studio running in the background of everything else.

Evening

The content factory runsagent

Episodes queue and render — a 120+ episode audio drama, 75 motorsports videos, and counting. A media company, staffed by agents.

Overnight

Monitoring never sleepsagent

Sixteen monitors watch uptime, content freshness, SEO, and security across every property. If something breaks, I hear about it — often after it’s already fixed.

I show up in about four places on this clock. Agents run the other twenty hours. That’s not a trick — it’s a structure, and it’s the same structure I install for clients.

How it’s organized

My org chart has one human on it.

Every function that would normally be a hire is a department of agents instead. I’m the only person — and my job is judgment, not production.

Peter — Operatorthe only human · sets direction, approves, decides

Consulting

Discovery, process maps, ROI ledgers, and the engagement dashboards clients actually read.

Web Studio

Builds and maintains client sites — draft, verify, ship, report — on a monthly loop.

Content

Two videos a day plus a standing library of episodes, all scripted, voiced, and rendered by agents.

Investing

Research, monitoring, and the discipline of testing a thesis before a dollar moves.

Operations

Backups, self-healing, monitoring, and reporting — the back office that runs itself.

This is a real org chart, not a slide. Build your own with the same tool I use — name a head agent per department and see the whole team on one screen.

The honest ledger

What it costs to run all of this.

People assume this is expensive. It isn’t — because there’s a difference between running a system and building one.

Under $400 / month

To run the whole operation — 9 sites, 36 automations, the daily video factory, 1.18M pages, all the monitoring. Less than the day-rate of a single junior hire. Once it’s built, it just runs.

~ $2,000 / month

Separate line, told honestly: lab work — building new agents and heavy experimentation — costs more. Building always costs more than running. But you build once, and then it’s a fixed asset that pays rent.

The one I’m proudest of losing

I built an AI that trades. Then I killed it — in public.

Over four months my agents ran 18 numbered experiments across two engines, scanned 19,072 markets autonomously, and tried to make money as a market-maker. Here is the exact moment the idea died — the same strategy, three numbers, each one a layer of wishful thinking stripped away:

If every quote filled
+$814
Theoretical edge — the number a spreadsheet shows you.
Top of book, real volume
+$149
Hypothetical — only 18% of the fantasy survives contact with reality.
Actual paper result
−$405
Honest — adverse selection fills only the losing leg. No edge.

Zero real dollars, ever

Every number above is paper. I pre-registered the decision gates, and when the honest layer failed all of them, I shut it off — without ever risking a cent of real money.

I published the autopsy

The full teardown — the mechanism, the mistake, the arc — is live on my own site. I don’t hide the failures. The discipline that kills a bad idea early is the same discipline I bring to your AI program.

This is what “test before you trust” actually looks like. The verifier, the gates, the kill switch — it’s the same rigor behind every client engagement.

Why this isn’t my first rodeo

I’ve built and exited before. Now I do it with a fraction of the people.

3 funds raised 5 exits 41 apps built with AI 27 Main-Street sites, agent-maintained 5 AI engagements in 12 months 10+ dev teams enabled

The point isn’t the tally. It’s that the person teaching you this has operated at scale the traditional way — and now runs more surface area than ever, solo, because the leverage is real. The full track record is on LinkedIn.

Want to run yours like this?

You don’t need a team — you need the structure. I teach it live, hands-on, and I’ll help you build your first agents in an afternoon. Or if you’d rather I build it with you, that’s what the engagements are for.

Learn to run yours → Or build it with me

Curious where you stand today? Score your own AI-enablement →