We are an AI-first product studio that builds custom AI systems to improve margins, free up capital, and reduce operational risk.
We work with SMEs to identify where time, capital, and operational risk are being lost, then build AI systems around those workflows and help teams develop the capability to use them well.
We believe the strongest AI systems are built around practical improvements that compound over time.
The phrase 'marginal gains' was popularised by Sir Dave Brailsford during his time leading British Cycling. His philosophy was simple: improve many small things by 1%, and the combined effect becomes transformational over time.
We think the same principle applies to AI inside businesses.
Most businesses don't need a dramatic AI rollout on day one. They benefit more from focused systems built around the operational work that keeps slowing the business down.
That might mean automating a repetitive process, building a proposal workflow, improving reporting, or giving agents access to the knowledge they need to do reliable work.
Each system on its own can seem contained. But once those systems are running, the effect compounds: margins improve, capital gets freed up, and operational risk starts to come down.
That is the thinking behind our Custom AI Systems work, Speedrun, and the longer-term support relationships we run with businesses that want to keep building capability over time.
Our work is built around a simple commercial framework: up, up, down.
We build systems around the parts of a business where the impact is clearest: grow margins, free up capital, and reduce risk. Capability-building still matters, but it works best when teams are learning around systems already solving real operational work.
Grow margins
We build AI systems around workflows where repetitive manual work is draining time from the people who should be serving clients, delivering work, and moving the business forward.
Free up capital
Better visibility, faster quoting, tighter inventory management, and cleaner financial workflows all help release capital that would otherwise stay tied up in slow, fragmented operations.
Reduce risk
Systems become less dependent on one person remembering everything, less vulnerable to missed handoffs, and less exposed to knowledge walking out the door when people leave or roles change.
Build capability
We run Speedrun Workshops, support longer-term development through Marginal Gains, and use Ambulando as the operating environment. That gives teams a practical way to learn, operate, and keep improving.
Our work is grounded in lived experience building and operating AI systems.
We help business leaders understand what AI can actually do by building real systems with it, not by treating it as a strategy exercise disconnected from operations.
We've worked with Stakwork, a Silicon Valley company building agent-based systems, and that experience shapes how we approach operational AI inside organisations. We care about what runs in the business, not just what sounds impressive in a workshop deck.
We operate as an AI-first business ourselves, using these tools across our own operations and building systems like Ambulando, our open-source agent workspace. We also continue to run workshops and education programs, including Speedrun for business teams and Level Up for schools, because capability develops fastest when people build with the tools directly.
Pete Winn
Co-FounderPete has a long track record in process redesign, deep tech and large enterprise and programme deployments from Rolls Royce to Rio Tinto.
Let's talk about what's slowing your business down.
Book a free intro call. We'll look at where time is going and whether a custom AI system makes sense for your business.
Book a Call