Services
Our work spans medtech, sports, web product, and AI consulting at the centre. It looks like four practices on a website; behind the scenes it's the same engineering judgment applied to four different problem shapes.
01 · Practice
Strategy, architecture, and shipping for AI projects from first evaluation through to production.
What it is
We bring senior engineering judgment to AI work — brought in to do the data analysis that frames the problem, design what's going to be built, choose what stays in scope, and stay close enough to the code to make the architecture actually hold. The goal isn't a deck. It's a system that's still running six months after the engagement ends.
Where it tends to fit
Teams who've prototyped an AI feature and now need it to be reliable. Companies whose first attempt was outsourced and didn't land. Founders who need a technical lead for the AI side without hiring a full-time CTO.
Activities
02 · Practice
AI engineering across the medical device space — inside the devices themselves and through the factories that build them. The regulated end of the field, where reliability and traceability are the floor.
What it is
AI and software engineering for medical device companies — both inside the devices themselves and across the manufacturing systems that produce them. Inside the device: production AI features, structured reasoning where it earns the verification cost. In the factory: Factory 4.0 / industrial connectivity, legacy machinery joined to a modern Unified Namespace, real-time operational dashboards for the floor. The discipline is the same in both: every change has an audit trail, every system is verified, the regulatory frame shapes design from the first sketch.
Where it tends to fit
Medtech companies adding AI to existing devices or workflows. Medical device manufacturers connecting legacy factory floor equipment (PLCs, OPC UA, MQTT) to modern data architectures, or putting AI into production-line operations. Healthcare-adjacent SaaS that has to behave like a regulated tool. Teams who have a software-of-medical-device or operations-of-medical-device problem and need engineering that takes the constraints seriously.
Activities
03 · Practice
Modern web applications with AI properly inside. Built for teams who will own the code after we hand over.
What it is
Full-stack product work where AI is one of the ingredients rather than the entire pitch. Next.js, TypeScript end-to-end, Payload or similar headless CMS, a real database, server-rendered where it should be, client-side where it has to be. Code your team can read.
Where it tends to fit
Companies replacing a tired existing product with something that needs to last another decade. Founders who need senior engineering to start the codebase well before a team scales it. Teams who've been burned by hourly agency work and want a small studio that builds the way they intend to maintain it.
Activities
04 · Practice
Software for clubs, federations, and organisations across the game — from membership to analytics to athlete-facing tools.
What it is
The sports work sits at the intersection of community technology and applied AI. Membership platforms, fixture and clubhouse operations, performance analytics. Lighter regulatory frame than medtech, but no less real — the people on the other end are members, parents, volunteers, athletes, coaches.
Where it tends to fit
Clubs whose existing tooling is a stitched-together pile of spreadsheets and free-tier SaaS. Federations modernising their member experience. Sports tech startups who want senior engineers who actually understand the domain rather than just the buzzwords.
Activities
Beyond the tracks
Some work doesn't sit cleanly inside a track. Two shapes come up enough to be worth naming.
Programme
Short engagements — half a day to a week — for teams that are adopting or scaling AI and want to do it from a position of informed engineering. Practitioner workshops for engineers (architecture, evaluation, common pitfalls); executive briefings for leadership (what's real, what's hype, how to buy). Tailored to the problem, not slide decks.
Programme
Modernisation programmes for organisations whose existing systems are holding them back. Replacing legacy platforms while keeping the lights on, integrating AI without ripping out what works, training the team to own what we hand over. Months-long engagements that often touch multiple tracks at once.
Work with the studio
Bring the problem, not a brief. The first conversation is about whether it's a fit — for both sides.