Why I Built Booksmaster

I've started about 50-60 businesses over the past 20 years. Most failed. Some limped along for a few months before I pivoted to the next idea. A couple had minor success that just didn't sustain.

That's not a confession. That's just entrepreneurship.

The last three years alone: I wrote a book of personal philosophy (published one copy, got zero feedback, moved on). Tried selling solar door-to-door for 9 months (sold 3 systems to friends/family, hated cold sales). Did AI consulting through the Chamber of Commerce (one unpaid consultation, never heard back—not even a thank you). Launched AI-generated graphic t-shirts on Etsy for 3 months (sold one unit online, gave away more than I sold, still have half the inventory in my garage). Created 10 AI short films on YouTube (first one hit 15k views, couldn't replicate it despite vastly better production). Built an X/Twitter following for 6 months (decent growth, not fulfilling). Finally pivoted to AI-assisted software development in February 2025.

That last one stuck. For now, at least. It's been almost a year—the longest I've committed to anything in this recent cycle.

But here's the thing: through all of those ventures, one pain point never changed.

The Receipt Problem

In 2012, I was running an eBay sporting goods store with a brick-and-mortar storefront. I did about 4,000-5,000 individual sales that year. I had somewhere around 400-500 receipts and a hundred more supporting documents to track—cost of goods sold, rent, insurance, local sales, inventory purchases.

Tax season rolled around. I looked at QuickBooks. Spent one afternoon trying to figure it out.

I gave up in a day.

Not because I'm lazy. Because QuickBooks wanted me to take a college-level course to learn their software so I could pay them $30-50/month for features I'd never use. Their receipt scanning? Tesseract OCR from 2010 that couldn't tell a scale ticket from a dinner receipt.

So I did what a lot of perpetual entrepreneurs do: I built my own solution.

If I hadn't kept up with it monthly? Easily 80-100 hours. Or a couple thousand dollars to pay someone else to deal with my shoebox of chaos.

Fast Forward to 2026

A friend—a general contractor—knew I was doing AI-assisted development. He asked if I could help him with his books. He'd been running his business for a year and had a mountain of unsorted receipts.

I thought: "I remember this nightmare."

I also thought: "AI vision models are pretty good now. Maybe I can build something better than what existed in 2012."

I didn't know if it would work. But once I started building, I realized something:

The tools still haven't caught up to the need.

Xero, Wave, QuickBooks—they're all still using the same garbage OCR they've been using for over a decade. They call it "intelligent" or "smart" or "automated." It's not. It's pattern matching and character recognition. It fails on manifests, it can't link payment receipts to itemized receipts, it doesn't understand context or graphic logos.

Meanwhile, they're selling you payroll features, inventory management, multi-currency support, industry-specific templates—none of which matter if you can't get your damn receipts sorted.

What I Built Instead

Booksmaster uses Gemini AI—actual artificial intelligence that understands context, not just characters.

It sees "Inbound Weight: 8860 lb" and knows that's a manifest, not an expense. It links your payment receipt to the itemized receipt automatically. It categorizes and learns based on patterns in your business, not just keywords. It gets smarter the more you use it.

And it works 6x more accurately than OCR, 5x faster. I didn't expect that. I knew AI would be better, but I didn't know it would be that good.

Why This Matters

Here's what Xero and QuickBooks won't tell you:

• • 20% of businesses fail in year one.
• • 50% fail by year five.
• • Most entrepreneurs try 3-4 ventures before finding success. (I'm at 50-60 and counting.)

The big accounting software companies are built for businesses that scale—hire employees, need payroll, manage inventory, expand to multiple locations. That's not most of us.

Most of us are trying idea #3 after idea #2 didn't work out. We're solo contractors doing good work but bad bookkeeping. We're freelancers who just need receipts categorized without a six-month learning curve.

We're perpetual entrepreneurs. We pivot, we try again, we start over. That's not failure—that's the process.

And through all of it, the receipts keep piling up.

What I Learned in 20 Years

I've been chasing entrepreneurship since high school. Four years in the military, then 20 years of trying to build something of value. I've had formal jobs to pay the bills, but I've never stopped trying to make my own thing work.

Why keep going after 50-60 failures?

Because I'd rather believe in myself than spend my life realizing someone else's dreams. Because some of those "failures" taught me more than any success could have. Because the Venn diagram of having a product, having a market, and having the timing align is brutally difficult—but it's worth trying.

If you're on business idea #3 and losing hope, here's what I'd tell you:

The next one might fail too, but you'll do better. You'll learn more. Eventually, you'll find your fit.

And even if you don't? There's nothing I'd rather be doing than trying.

Why Booksmaster Exists

I'm not building this to sell you a fantasy about scaling to 100 employees. I'm not assuming you'll outgrow me in 3 years and graduate to QuickBooks.

I'm building this because I know what it's like to be stuck in the perpetual entrepreneur cycle. I know the dread of tax season when receipts are a mess. I know the frustration of software that offers everything except what you actually need.

If you're a contractor, a freelancer, a serial entrepreneur on venture #4, a solo business owner who just wants accounting to not suck—I built this for you.

Because I am you. And I needed this 10 years ago when I was drowning in 400+ receipts with nothing but a calculator, a custom spreadsheet, and a dream.

Now it exists. And it actually works!