INITIALIZING SYSTEMS
Trevor

I BUILT
A THING

I fix problems by building systems.
Some work. Some don't. This is the record.

PROBLEM → BUILD → TEST → REPEAT
SCROLL TO EXPLORE

THE
ARCHIVE

A living record of reality.
Swipe or scroll →

01
AT CAPACITY

Your Free CRM

Every CRM I used forced me to adjust my workflow to fit the tool.

  • → Built a fully tailorable CRM system
  • → Users can turn off features they don't need
  • → Opened as a free beta for testing

3,500 users. 88% in it daily. At capacity.

02
ACTIVE

Chore Bucks

Got tired of reminding my kid to do chores.

  • → Built a system that ties chores to pay
  • → Do it → get paid
  • → Don't → nothing

Still used daily.

03
INTERNAL

CRM + Dialer + QA System

Outbound calls were blind and useless.

  • → Built a system that shows real website data during calls
  • → Added guided prompts based on actual issues
  • → Analyzes performance automatically after

Internal tool. Actively used.

04
FREE

Focus AI Website Scanner

Got tired of running 5 tools to analyze one site.

  • → Built a scanner that does it all in one place
  • → Pulls real data from Google APIs
  • → Checks 155 different factors instantly

155 checks. Pulls real data from Google. Free.

05
TESTING

Klear Notes

Got tired of messy notes and trying to organize thoughts manually.

  • → Built an AI tool that organizes thoughts automatically
  • → Input raw lists → it categorizes by Tasks, Shopping, Notes
  • → No formatting or thinking required

Used daily. Still in testing.

01

Your Free CRM

WHY I BUILT IT

For as long as I can remember, I've wanted a CRM that actually fit how I work.

Every one I've ever used felt the same. It's a one size fits all system where you're expected to figure out how to make it work for you. There's hundreds of features you don't need, menus everywhere, and half the time it just slows you down instead of helping.

It always felt backwards. Instead of the tool supporting your workflow, you end up adjusting your workflow to fit the tool. That's where the friction comes from.

I didn't want something bloated. I didn't want something I had to "learn." I wanted something that could be tailored to exactly what I needed, nothing more, nothing less.

So I built it.

WHAT I BUILT

A CRM that's fully tailorable.

You can go into settings and choose what you want to see and what you don't. If you want pipelines, use them. If you don't, turn them off.

Everything is optional.

The whole point is simple. The system adapts to you, not the other way around.

I opened it up as a free beta and let people in.

WHAT HAPPENED

About 3,500 people got access.

It's been a few months now and it's still actively used. About 88% of users are in it daily.

That's the part that matters. Not signups. Usage.

People don't bounce off it because it doesn't get in their way. A lot of them just stay on it and keep using it.

WHAT DIDN'T WORK

It was never built to scale beyond this.

Right now it's close to the capacity of the server it's on. It still runs smooth, but opening it up to more people would require infrastructure changes.

At that point it stops being what it was meant to be and turns into something else.

That's not really the goal.

CURRENT STATE

At capacity.

Still running clean. Still used daily.

I use it. Other people use it.

I'm not actively trying to expand it. I'm honestly pretty content with where it's at. It solved the problem it was built for.

02

Chore Bucks

WHY I BUILT IT

I got tired of reminding my kid to do basic chores.

Every day it was the same cycle. If I didn't say anything, nothing got done. If I did, it turned into me managing the entire process instead of her actually being responsible for it.

That's not a system. That's babysitting tasks.

I wanted something where the expectation was clear and the outcome handled itself. No reminding, no chasing, no back and forth. Either it gets done or it doesn't, and the result follows that.

I also wanted her to start understanding money a bit better. Not just earning it, but actually seeing it go somewhere when it's spent.

So instead of repeating the same conversation every day, I built something that handled it.

WHAT I BUILT

A simple app to track chores and tie them directly to daily allowance.

She logs in and marks her chores as done. I verify them. If everything's completed, she gets paid for the day. If not, she doesn't.

No reminders. No notifications. No system chasing her or me.

Also built in:
Spend tracking so she can log what she buys
A running balance
Some basic consequence variables

The logo was thrown together in about 30 seconds with a ChatGPT prompt. Didn't even bother cleaning it up or vectorizing it. This was never meant to be polished, just functional.

WHAT HAPPENED

She started doing her chores without being asked.

That was the whole goal.

The system removed me from the loop. The rules are obvious, and the outcome is predictable. There's no arguing about it because it's not me deciding in the moment.

The spending tracker ended up being more useful than expected. Now she sees how quickly money disappears when she uses it, not just when she earns it.

WHAT DIDN'T WORK

Everything relies on manual verification from me.

There's no automation, no reminders, no safeguards. If I don't check it, it doesn't move.

UI is rough. The whole thing was built fast and it shows. Not designed to scale or be anything beyond this use case.

CURRENT STATE

Still used daily.

Built a few months ago and it's still part of the routine. Shared it with a couple parent friends and they're using it the same way.

It's simple, but it does exactly what it was supposed to do.

03

CRM + Dialer + QA System

WHY I BUILT IT

We launched a new product called Foundation. Simple offer. $1,000 site, $100 a month.

To move it, we started building out outbound.

I've never liked sales calls. Don't like getting them. Don't like making them. Most of them are garbage because the person on the phone has no real information. They're just reading generic lines hoping something sticks.

I didn't want that.

At the same time, we were looking at using a call center. I've got clients in that space, so I reached out to one I've worked with the longest. He walked me through the costs, and part of that stack was CRM, dialer, and all the usual tools.

The problem is none of those tools do what I actually needed.

I wanted the person making the call to see, in real time, exactly what that business's website is doing. Real data. Real performance. Something they can actually use in the conversation instead of guessing.

That doesn't exist in third party CRMs.

So instead of forcing it, I built it.

WHAT I BUILT

Started with an internal CRM.

As soon as a lead is added, it automatically scans their website and pulls a full report. Performance, structure, everything.

Now when the outbound agent calls, they're not blind. They're looking at actual data about that business.

On top of that:
The system shows guided talking points tied to the report
Not generic scripts, but contextual prompts based on what's wrong

Then we needed a dialer.
So instead of integrating one, we built it into the system.
Then we added call recording.
Then QA.

Now the full flow looks like this:
Agent makes the call →
Sees real-time website data →
Uses guided prompts tied to that data →
Call is recorded →
Recording goes to AI →
AI analyzes performance, quality, gaps →
Second AI turns that into a clean report for management

So now we're not guessing where calls go wrong. We see it.

WHAT HAPPENED

Agents have real context during calls.

They're not just saying random stuff. They're referencing actual issues with the business they're calling.

Call quality becomes measurable instead of subjective.

Instead of listening to hours of calls manually, the system surfaces what matters.

It turned outbound from something I hate into something that at least makes sense.

WHAT DIDN'T WORK

This was never meant to be clean or simple.

It's layered, complex, and very specific to how we operate.

Not something that drops into another business easily.

Also not something I have any interest in turning into a product.

CURRENT STATE

Internal only.

Used for outbound tied to Foundation.

Still evolving, but fully functional.

No plans to release it. No real reason to.

It does exactly what we needed it to do.

04

Focus AI Website Scanner

WHY I BUILT IT

For years I was using a mix of tools to analyze websites.

Google PageSpeed. Google Rich Results. A bunch of third party scanners.

Every time I wanted a real picture of a site, I had to open multiple tabs, paste the same URL into each one, run all the scans, then piece the results together myself.

On top of that, I was paying for several of those tools monthly.

It was slow, clumsy, and honestly just annoying to deal with.

The bigger problem was the fragmentation. Each tool showed part of the picture, but none of them gave a complete view of how a site was actually performing.

I haven't found anything better that doesn't require multiple scans.

So I built something that did it all in one place.

WHAT I BUILT

A website scanner that runs a full audit across performance, structure, SEO, accessibility, and infrastructure.

It pulls real data directly from Google APIs like PageSpeed and Rich Results, then layers additional analysis by scanning the site itself.

Right now it checks around 155 different factors, including:

Core performance metrics like LCP, TTI, CLS, TTFB
Image optimization and media handling
Accessibility basics
Security headers and HTTPS setup
SEO signals and structured data
Document structure and heading hierarchy
Crawlability, indexing, and sitemap integrity
Internal linking and navigation
Mobile usability
Server and infrastructure configuration

There's also weighting behind it.

Not everything matters equally, and that changes over time. When Google shifts what it cares about, I adjust the scoring and priorities inside the scanner.

It's not static.

WHAT HAPPENED

It removed the need for multiple tools completely.

Instead of running five different scans and trying to combine them, everything shows up in one report.

It's faster, cleaner, and gives a more complete picture of what's actually going on.

Because it's pulling directly from Google where possible, the data reflects real-world signals, not just assumptions.

WHAT DIDN'T WORK

It's not perfect.

The weighting isn't always right and gets adjusted over time.
Sometimes things show up slightly off and need tuning.

Also, this level of depth can be overkill for someone who just wants a quick answer.

CURRENT STATE

Live and free to use.

No signup. No email required.

I built it to be used, not gated.

It's gone through a major overhaul, expanding from around 80 checks to 155.

Still evolving as standards change.

05

Klear Notes

WHY I BUILT IT

I got tired of trying to keep everything in my head or scattered across random notes.

Half the time it's just a brain dump. "I need to do this, this, this… pick this up… don't forget that…"

Most note apps don't handle that well. You either structure everything manually or it just turns into a mess you don't go back to.

I didn't want to organize notes. I wanted to just say what I needed and have it make sense automatically.

So I built something that does that.

WHAT I BUILT

An AI note taker that organizes things as you dump them in.

You can throw in a messy list like:
"I need to do this, buy this, don't forget that…"

And it sorts everything into:
Tasks
Shopping lists
General notes

No formatting. No thinking about structure.

Just input → organized output.

WHAT HAPPENED

I use it every day.

It removes the friction of trying to organize thoughts while you're having them.

Instead of stopping to structure things, you just get it out and let the system handle it.

A few friends are using it too, same way.

WHAT DIDN'T WORK

It's not fully pushed through Google's requirements yet.

So it's not a polished, public app sitting in the store.

Also not really built with monetization in mind.

CURRENT STATE

Functional and used daily.

Technically available, but still in testing stages.

Might take it further and push it through Google properly at some point… might not.

Right now it does exactly what I need it to do.