Work

What we have built, and what it does every day.

Each entry is a real system that is running today: the business situation, what we built, and what changed. We describe businesses by industry and never by name, and client screens show sample data.

Every number on this page comes from the live system, with its source and the date we read it.

For businesses

A multi-state life insurance agency

Inside it: Lead Response System, Call-Now Dispatch, pre-call briefs, the call log, Ads Shadow Agent, Daily Google Numbers

Running daily

A call list that puts the longest wait first.

The situation
People asked for help through the website and ads, and some waited hours for a call. The agent had no single place to see who was waiting, what they wanted, or what had already been said to them.
What we built
A command center the agent opens every morning, with twelve screens from today's calls to money. An alert goes out the moment someone asks for help, a written brief arrives before each booked call, and every call result is one tap. Proposed ad changes are checked with Google before anything is applied and scored 28 days later.
What changed
Every new lead now sets off an alert within seconds, every call result lands in one log with a receipt, and ad, search and website numbers arrive in the same system every day.

Under 2 secondsfrom a new lead arriving to the alert going out, for every lead in the last 45 days.

Source: the agency's lead system, read 30 October 2026

One person's page: what they asked for in their own words, their details, and a large call button
One person's page, shown with sample data. Their own words come first.
The call list: people waiting for a call, how long they have waited, their local time and buttons to record the result
The morning call list, shown with sample data.

Two self-storage facilities

Inside it: local search ads, Facility Market Watch

Running

Local ads and a twice-weekly read of the market.

The situation
Two independent storage facilities compete with national chains on Google, without a marketing department, and prices move around them every week.
What we built
Google search ads for each store, split by what renters look for: storage units close by, RV and boat parking from farther out, and truck rental. Facility Market Watch reads every nearby competitor's prices and recommends a price for each unit size.
What changed
The owner sees the local market before he sets a price, and every price stays his decision.

6search campaigns live across the two stores. Competitor prices are read every Tuesday and Friday.

Source: the stores' Google Ads accounts and our market watch, read 30 October 2026

No screenshot: these reports carry the owner's own prices.

A sports-nutrition shop

Inside it: Site Monitor, Daily Google Numbers

Running

A website someone is watching.

The situation
A local shop relies on its website to bring people through the door, and needs to know the site is up and being found.
What we built
Site Monitor checks the site around the clock and audits its search basics every Monday. Daily Google Numbers brings its Google visit counts into our system every day, next to every other site we run.
What changed
A broken page or a drop in visits shows up for us first, not for a customer at the door.

Every 15 minutesSite Monitor checks each of the 9 websites we run, this shop's included.

Source: our monitoring system, read 30 October 2026

No screenshot: the reports show the shop's own numbers.

A family energy company

Running

The system a family business runs its records on.

We built, and keep running, the records and reporting system a family-owned energy company uses. That work stays private.

Our own products

We build products of our own, and they are where we try things first.

The AI Roadmap

A product for business owners

On sale

A personal AI plan any owner can buy online.

The situation
Owners hear about AI everywhere and cannot tell which of it is worth their money.
What we built
A twenty-minute intake, a written plan of 8 to 12 ranked moves with tools, costs and a first 30 days, secure checkout, and a private workspace where the owner tracks progress, asks questions and requests a quote. Caleb reviews every plan before delivery.
What you can do today
Read the complete fictional sample, then buy your own at the founding price.

See the AI Roadmap   Read the sample

The Roadmap workspace for a fictional insurance agency: the next move, plan progress and a scorecard
The buyer's workspace, from the public fictional sample.

Noval OS

Our own operating system

In daily use

How a small company runs many businesses without losing track.

The situation
When one team runs work for several businesses at once, decisions, to-dos and research disappear into email and memory.
What we built
Noval OS keeps every project, to-do and decision in one place, with the reason for each decision written down. A searchable library answers questions from our own research, and a model trained on what we actually finished ranks what to do next.
What changed
Nothing important lives only in someone's head, and the model's picks are checked against what really happened.

1,661decisions recorded with their reasons since April 2026.

Source: Noval OS, read 30 October 2026

No screenshot: it holds client records.

Roblox games

Games published under the Noval Studios group

Two public

Games built with AI tools, played by real people.

The question
Can a small team use AI to design, build and publish games that players choose?
What we built
Two public games: Pets Eat Monsters! and Grow a Giant Snowman!, which launched in the fall as Grow a Giant Pumpkin!
What we learned
Across every game we built, fixing each wrong screen by hand cost days. Fixing the tool that made the mistake fixed every copy of it at once, and that is how we fix things for clients too.
Pets Eat Monsters! game art: food pets swarming a giant vegetable monster Grow a Giant Pumpkin! game art: players carrying pumpkins to a weighing scale at a farm
Game art from the two public games.

Want one of these for your business?

Start with a free check-up. You get a one-page snapshot of the top three places AI would help, and 20 minutes with Caleb.

Or see every way to work with us.