← The Automation Files
How We Work 7 min read · April 2026

You have the skills. I’ll sell them.

How Onizuka Studio actually started — not with a business plan, not with investors, not with a clean origin story. With a paralegal, a developer, a baby on the way, and a bet on complementary skills.

Michael had been building DNN websites since version 4. That’s not a small thing — DNN is a serious enterprise CMS, and being fluent in it at that level takes years. He was good. He was genuinely, demonstrably good.

He was also, by his own honest assessment, not a sales person. Not a client communication person. Not the kind of person who picks up the phone and makes a potential client feel immediately taken care of. Those skills are real and specific and not everyone has them — and there’s nothing wrong with that. It just meant there was a gap.

I was the gap.

We found out we were expecting. I made the call.

I was working as a paralegal. I was good at it — Dean’s list, ran the student association, the whole thing. But when we found out we were expecting, we looked at what we had and made a decision. Michael had been doing freelance web work. There was something real there. We just needed someone to run the business side of it properly.

So I told him: you have the skills, I’ll sell them. I’ll learn what I need to along the way.

That’s the whole founding story. No pitch deck. No seed round. No carefully researched market analysis. A developer who could build anything and a paralegal who knew how to talk to people, make them feel heard, manage the process, and deliver what she promised.

We weren’t betting on confidence without substance. We were betting on a real combination of skills that genuinely complemented each other. We wanted to work from home and build something of our own. So we did.

What I brought that had nothing to do with websites.

I didn’t know what a website was. Not really. I understood the concept, obviously — but the technical underpinnings, the CMS architecture, the difference between a module and a skin? None of it. I was learning in real time, project by project, question by question.

What I did know: how to read a client. How to make someone feel like their project mattered and wasn’t going to get dropped between the cracks. How to translate what they were trying to say into something a developer could actually build from. How to set expectations clearly so that "on time and on budget" wasn’t a happy accident but a system.

I also knew — from years in law firms — how to set up an operation. Before we took on our first real client, I had the project management system built, the file structure set up, the invoicing running, the client communication templates drafted. I wasn’t going to run a technology company on sticky notes and good intentions.

Clients noticed. Almost immediately, we started hearing versions of the same question: how do you have all this figured out? Can you do that for my business?

Then clients started asking about the other thing.

In every project, at some point, clients would start venting. Not about the website — about everything behind it. The chaos nobody had organized. The process that took four people to execute that should have taken one. The tool they’d bought that nobody had set up correctly. The spreadsheet that was basically held together with duct tape and institutional memory.

I couldn’t walk past that. I’d start making suggestions. They’d start taking me up on them. So I started building internal systems — Zoho CRM setups, workflow automations, document pipelines, the whole thing. One client’s Zoho turned into five clients’ Zoho. Five turned into custom-built apps. The automation work eventually grew larger than the website work.

Same skill. Different application. Read the problem, build the solution, make it run without someone having to babysit it.

What this means now, with AI in the picture.

Michael is still the coder. When something needs to be built from scratch — a custom module, a DNN skin, a piece of functionality that doesn’t exist off the shelf — that’s Michael. That skill is deep and real and not easily replaced.

AI is mine. Not because I sat down and decided to learn it — because it clicked in a way that felt immediate and different from any other tool I’d picked up before. It works the way my brain works. Pattern recognition, systems logic, taking what someone describes and turning it into something structured and functional. When I started working with AI seriously, I didn’t feel like I was learning a new tool. I felt like I’d found a better version of something I’d been doing manually for years.

So the studio has always had exactly what the work required. The skills just kept growing to meet what was needed. In 2013 it was Michael’s DNN expertise plus my client and operations skills. In 2018 it was Zoho and workflow automation. Now it’s AI implementation and custom-built systems that didn’t exist five years ago.

We’ve never been in over our heads. We’ve just kept building on the same foundation: know your skills, be honest about what you don’t know yet, and don’t take on work you can’t deliver. That hasn’t changed since day one.

Michelle Onizuka

Co-founder of Onizuka Studio. She runs discovery, client relationships, systems design, and automation builds. Michael runs the code. Together they’ve been building digital operations for small businesses since 2013. The full story →

More from The Automation Files

How I Work

I don’t sell features. I sell the why.

AI & Automation

AI fits the way my brain works. Here’s what that actually means.

Tell me where you are. I’ll take it from there.

Pen and paper or already on Salesforce — it genuinely doesn’t matter. I’ve started from worse.

Start the conversation →