Who builds GattyWorks

The question we get most is a polite version of: how is that even possible with that few people? It is a fair thing to ask. The answer is not a clever process or a string of late nights. It is who is on the team and what they have built before, so that is what this post is about.

We are a senior team of designers and engineers, plus an AI agent fleet we treat as part of the team, not a toolbar. Most briefs are touched by the same hands from the first reply to the deploy.

The track record

The engineering came first. People on the team have spent more than six years shipping production software, well before ChatGPT made AI a headline.

That includes computer-vision work in the .NET era, training and testing models against ImageNet, the benchmark image dataset most recognition models were measured on.

It includes ML research with a university lab on noisy image data. The experiment was easy to state and hard to do: add controlled noise to an image, have a model try to reconstruct and recognize the original, and report a confidence score for how sure it was. You learn a lot about a model by watching where its confidence falls apart.

It also includes the unglamorous backend work that real products run on. Async APIs in Python and Go, sized for hundreds of thousands of tasks a day. Distributed data across Postgres, MongoDB, Redis, and search clusters. Payments and subscriptions, auth and single sign-on, offline sync for mobile apps, and end-to-end encryption that lets clients upgrade without forced app updates. Some of these systems handle more than a million requests a day.

And it includes the layer below the app. People on the team have written firmware and built cloud platforms for hardware and smart devices: vision systems, video streaming per device, and IoT fleets talking over MQTT. Web, mobile, and the hardware itself.

Earlier work runs from robotics to a stack of custom websites and apps for a college. Different domains, one habit: build the real thing, measure it, fix what breaks.

More than engineers

A studio that only writes code ships products that are hard to sell. So the team is mixed on purpose.

• A product lead with an MBA and real marketing and advertising experience, who thinks about the business case, the message, and the layout in the same breath.
• Recruiting and talent experience, which is how we pull in the right specialist when a brief needs one.
• Engineers across web, mobile, data, AI, and the firmware that runs on hardware, who have shipped to production, not just to a demo.

People from different domains, working with AI and agents, is the whole point. The agents do the repetitive heavy lifting. The humans make the calls that need taste and judgment.

What we build for

We build for profit and AI-forward companies: production-grade software with tests, monitoring, and a secured knowledge base, not a weekend prototype.

It also means we are careful about what we take on. We take a limited number of briefs at a time so the senior team actually stays on your project. That is the honest tradeoff. We are not the cheapest, and we are not infinite. We are fast, and we are senior.

We build unique things, the kind that did not exist before the brief, and we treat clients as partners, not tickets. The work is better, and it lasts longer, when both sides want the same outcome: something real that people use, that brings them joy and makes an impact.

What is coming

We are going to publish what we learn. Cookbooks and playbooks from real sprints, the same notes we use internally. If you want to see how the 24-hour cadence holds up under a real brief, this blog is where that will live.

Next step: send us a paragraph about what you want to build. One email is enough. hello@gattyworks.com.