Blown the Bloody Doors Off: How AI is Rewriting the Rules of Wealth Tech Development
My journey into becoming a Claude aficionado began last week with two chance conversations. The first was with Nick Raine, the CEO of Soderberg & Partners Wealth Management. Nick showed me a client onboarding app they had developed using Claude in 10 minutes. The app was stunning, super easy to use and very easy on the eye. Nick and I produced an outline design for the onboarding journey about a year ago, and it went into the dev backlog, where it has been gathering dust. Claude has blown the doors off traditional development paradigms. It enables anyone with a product-owner mindset and industry experience to develop fully functional, industrial-strength apps – without being stymied by a lack of dev resources.
The second chance encounter was with another ex-colleague, Rob Duncan, with whom I first worked almost 15 years ago when Novia Financial partnered with Aegon to build their ARC platform. Over the intervening years, Rob and I have worked on several platform implementations, essentially serving as the glue between clients and suppliers and accelerating delivery through Scaled Agile methodologies and practices. Rob contacted me out of the blue to pitch his new AI-driven ecosystem - he said something that stopped me in my tracks.
“I've spent 20 years in the platform industry watching product owners fight for dev resources. I've sat in steering committees where a six-screen onboarding journey takes nine months and £300k to deliver. I've watched firms wait 18 months for a platform supplier to ship a feature that should have taken weeks”.
He claimed that the ecosystem he has built will give product owners superpowers, enabling them to use epics and user stories to produce scalable code with full documentation, testing, and deployment. Because Rob is the most gifted person I have ever worked with, and Nick Raines' warm-up earlier in the week, I was hooked. Then he showed me what he had built. Not a mock-up. Not a slide deck. A working, architected solution with the same layered documentation you'd expect from a tier-one platform provider — system context diagrams, container diagrams, component diagrams, code sequences. The full C4 model (the dev guys will know what I mean). The kind of output that satisfies enterprise architects, SecOps, and compliance. Production-grade code with audit trails, not a working prototype held together with duct tape.
By Easter Saturday, I had built my first app. A simple but very usable cashflow modelling tool designed to optimise drawdown investments across ISA/SIPPs and other income sources. Maximising income whilst minimising tax.
At Wealthtech, our core products and services focus on migrating from legacy to modern, API-enabled services, typically providing workstream expertise for multi-year, multi-million-pound projects. We have been able to automate migration, and now the critical path is development (usually but not always) from a third-party vendor, platform, CRM, data warehouse, or portal. As go-live deadlines approach, the pressure to deploy additional non-existent development resources becomes apparent. Back in the day, FNZ would bring in the cavalry from other projects to protect the deadline. What Rob showed me on Friday wasn't just faster development. It was a fundamentally different model for how financial services software gets built — and it has a name. Rob has developed an approach he calls Specification-Governed Development, or SGD. It brings together six established industry-standard disciplines — architectural decision records, behaviour-driven development, spec-driven workflows, design systems, capability modelling, and intelligent dependency tracking — into a single, unified framework that works hand in glove with AI. You don't need to understand the six disciplines to understand what SGD does. Here's what matters:
Every piece of code traces back to a specification that a product owner can read in plain English. No more "trust us, it works." The spec is the contract, and it's written in your language, not the developer's.
Architectural drift — the silent killer of large codebases — is eliminated. When you've got hundreds of repositories across multiple teams and suppliers, they stay coherent. The same patterns, the same standards, the same governance. Automatically.
The output passes enterprise inspection. The governance artefacts, the audit trails, have the same rigour you'd expect from a tier-one platform build — delivered in a fraction of the time and cost.
For platform CEOs/CTOs/Programme Directors, SGD means your existing dev teams ship faster and more consistently, without the need to hire or spin up additional sprint teams. For large wealth firms, it means you can commission bespoke tooling — onboarding journeys, suitability engines, migration utilities — at a price point and timeline that was previously impossible. For a CTO, it means the AI tools your teams are already experimenting with finally have the guardrails to produce production-grade output.
In the coming weeks, Rob and I will be unpacking how SGD works in practice, what it means for platform suppliers trying to keep pace with demand, and how it changes the economics of building wealth technology. But here's what I want you to take away today: the gap between what your firm needs to build and what you can afford to build just closed.
#WealthManagement #FinTech #DigitalTransformation #WealthTech #AIDrivenDevelopment