How did we get here?
We started in 2013 building websites. We didn’t plan to become an AI integration and automation studio. The work took us there — one client problem at a time. Here’s the actual story of how Onizuka Studio got from there to here.
2013. Websites, web apps, and the occasional “hey, have you tried this tool?”
We launched Onizuka Studio in 2013. Michael builds things — front-end development, DNN, custom client portals, web applications, databases. We built websites for wineries, agricultural businesses, shooting sports organizations, government clients. Things were good. Growing steadily, doing quality work, building a reputation.
Beyond the website work, clients would sometimes ask what we were using to run our own business, or what we thought about their project management setup, or whether we had any recommendations. And we did — because we’re the kind of people who genuinely get excited about a good tool. Zoho. Teamwork. Basecamp. Jira. We’d find something that worked well and immediately want to tell everyone about it. Sometimes that turned into “okay let me actually just help you set it up.”
That wasn’t a service line. It was just us being us. But it planted seeds we didn’t realize were growing.
2020. We became vampires.
While the world was at home on TikTok making sourdough and binge-watching Netflix, we were completely unreachable. Not because we weren’t there — because we never stopped working. Every business that had been operating face-to-face, in-person, on paper, suddenly needed to be online. Not sometime soon. Now. This week. Before Monday.
The phone did not stop ringing. Restaurants needed online ordering. Service businesses needed booking systems. Consultants needed client portals. Physical retail needed e-commerce. Medical offices needed telehealth. We went from website developers to digital emergency responders almost overnight. What that actually felt like from inside the work is a different story — but from a business perspective, it changed everything about what people needed from us.
We weren’t the nurses or the doctors or the delivery drivers. But we were the ones keeping every business behind them running. The ones who didn’t get thanked because nobody really understood what we did or how fast we were doing it. We just showed up and built whatever needed building until it worked.
We saw it happening in real time. Businesses building digital operations the way you’d pack a bag in an emergency — grabbing what’s closest, not what’s best. We helped where we could but the demand was bigger than any of us could keep up with. People built what they needed to survive. And they survived.
2022. The calls started changing.
By 2022 the nature of what people were calling us about had shifted. It was still growing — still more work than we could take on — but the character of the problems was different. Less “I need a website” and more “I need to make this process digital.” Less “what tools should I use?” and more “how do I get these tools to talk to each other?”
People were sitting on stacks of subscriptions that each did their job fine in isolation — and none of them were connected. Someone was manually copying customer information from the booking system into the invoicing system every day. Someone else was exporting an email list every week and re-uploading it somewhere else. Documents were being emailed back and forth and re-filed by hand. The operation worked, technically. But it only worked because a person was running between all the pieces holding it together.
Digitizing paper workflows. Connecting platforms that had never been designed to connect. Automating the repetitive manual steps that were eating hours every week. This started becoming the bulk of our work, and it stayed that way.
2026. The tools finally caught up.
We’ve been watching AI develop for years. Using it the way everyone uses it — research, discovery, drafting, exploring ideas. But we kept the actual building in-house, because for a long time the tools weren’t good enough to trust with client work. They could generate plausible-looking code that broke in production. They could suggest solutions without understanding the system they were supposed to fit into.
Something changed around January 2026. Real production-level work became possible. Not just “it can write a function” — complex integrations, multi-step automations with real business logic, systems that understood what they were supposed to accomplish. We tested it carefully. We ran it on client work. We validated that the quality held. And then we changed how we operate.
What that means for a small business owner wearing ten hats is significant. The bottleneck that used to be “finding the right developer who has time” or “this is too complex to build on our budget” — that bottleneck is smaller now. Things we could scope but couldn’t build affordably for a small business, we can build now. Things that would have taken months, we can scope and deliver in weeks.
The mess from 2020 — six years of subscriptions and patches and manual workarounds — is costing businesses real money. Almost a million dollars. That’s how much one business may have lost since 2020 in operational drag we uncovered in an audit this year. See the report → That’s fixable now in ways it wasn’t two years ago.
That’s how we got here. Not a pivot. Not a rebrand. Thirteen years of following the work to wherever the real problems were — and the real problems, right now, are the ones this site was built to solve.
If your business is still running on 2020-era patchwork — or if you’ve been sitting on pen and paper while competitors pull ahead — start with an Automation Audit. We map what’s actually costing you money and tell you exactly what it would take to fix it.