Jacob Bettany & Niche Clever

Publishing was the first part of my working life that really stuck. Before joining Institute of Physics Publishing in 1999, I worked with banking and country-reference information at Reed Business Information. At IoP I moved from publishing administration into business development, worked on the launch and development of Quantitative Finance, and later took publisher responsibility for the journal.

Looking back, publishing gave me habits I still use: paying attention to sources, understanding the audience, making complicated material navigable and taking responsibility for the process that gets something finished and out into the world. In 2004 I left IoP, set up Niche Clever and launched MoneyScience.

MoneyScience began as specialist publishing, curation and community around quantitative finance, but over the years it pulled me further into websites, databases, online communities, training and long-running digital systems. Client work did the same. Through Thematic and related platform work I spent years between clients, users and developers, working on requirements, configuration, testing, training and the practical business of keeping specialist platforms useful.

I often find that the thing I’ve been asked to work on depends on a wider system around it. A website may turn out to be a publishing problem. A platform problem may really be about how users describe their work to developers. An operations role can involve membership, payments, events, suppliers, communications, access and the knowledge people rely on without necessarily writing it down.

At IMLPO, a part-time Operations Manager role made many of those connections visible at once. London Drawing sits across events, video and web publishing. Work with Richard Craven began with books and increasingly involved keeping a large, interconnected literary and publishing estate usable. Trigger SMARTS began with a practical hotel-signage problem and led to me designing and building the software around it.

Those projects do not need to be versions of the same service. I tend to be most useful when I can get far enough into a particular situation to understand what is actually going on, especially when the boundary of the problem is not obvious at the start.

The big change from AI in my working life is recent. I had been interested in AI and technology long before generative models became useful day-to-day tools, but coding agents changed what I could do directly. For years I worked close to technical systems through requirements, configuration, testing and implementation, while some deeper engineering changes still depended on other people. That distance became more limiting later in my career.

Coding agents have altered it dramatically. I can now take a problem much further myself: understand the domain, work out how the information and workflow need to behave, build the interface and the logic behind it, test it, debug it and revise it against what actually happens in use. Trigger was the first project where I felt comfortable thinking of myself as a software developer; by the time I built Lettings Compliance, building useful software had become something I actively wanted to do more of.

I don’t pretend that this makes me a specialist in every technology I use, and there are still areas where deeper specialist engineering is the right answer. What has changed is the distance between understanding a problem and being able to make something working from that understanding. I am still working out exactly what that means for the shape of Niche Clever, but the practical change is already very clear in the work.