The honest case for AI in the enterprise
Most enterprise AI programmes fail for the same reason enterprise search failed in 2008: nobody owned the data. Here is what actually has to be true before a model earns a place in production.
Read essay →
Est. 1996 — 30 years in Information Technology
Transforming ideas into innovative digital solutions with cutting-edge technology. Specialising in enterprise applications, cloud architecture, AI, and robotics.
30+
Years in practice
04
Disciplines
Enterprise
Scale of work
Guides
AI · Robotics · Cloud
Math
Proofs that travel
The premise
On one side, the way work has always been done. On the other, the way it could be done. The engineering is never really about the technology — it is about carrying people across without dropping what they depend on.
I have spent 30 years on that crossing: from green screens and client–server, through enterprise Java and integration, into cloud, data platforms and now AI. This is where I write it down — the decisions that held, the ones that did not, and what I would tell a younger engineer standing at the near end of the bridge.
Core systems that carry a business for a decade: domain modelling, integration contracts and the unglamorous discipline of data that stays correct under load.
Landing zones across the major clouds, resilience and cost as design constraints — migrations judged by how much cheaper change became afterwards.
Language models, agents and physical systems put to work — separating the capability that compounds from the demo that does not survive contact with production.
Queueing, kinematics, statistics — the equations behind capacity, motion and evidence, written so they apply beyond any one stack or robot.
Four disciplines
Science asks why, mathematics proves it, engineering builds it, technology ships it. Everything here belongs to one of the four.
First principles
The habit of asking why before asking how — evidence, experiment and the discipline of being proven wrong.
The moving frontier
Cloud, data and artificial intelligence: what genuinely changes the work, and what is only noise in the feed.
Built to last
Architecture, trade-offs and delivery at enterprise scale — the craft of shipping systems that outlive their authors.
The formal core
Models, algorithms and the quiet structure underneath every system we claim to have invented.
Selected writing
Long-form essays on architecture, delivery and the technologies currently rearranging the profession.
Most enterprise AI programmes fail for the same reason enterprise search failed in 2008: nobody owned the data. Here is what actually has to be true before a model earns a place in production.
Read essay →Looking back across thirty years of systems, the choices I am still glad about have almost nothing in common technically — but they share a single quality: they kept the next decision cheap.
Read essay →Published in Clinical Rheumatology — what four years of annual biosimilar rituximab retreatment showed for disease control and safety in seropositive rheumatoid arthritis, and why real-world evidence still matters beside the trial literature.
Read essay →From the GPU in your cloud instance to the quantum chip that may one day break encryption — a map of 19 processor architectures shaping 2026 and beyond, with practical guidance on when each one actually matters.
Read essay →1996 to today
The tools were replaced roughly every seven years. The fundamentals never were.
1996 — 2002
Green screens giving way to three-tier applications, and financial modules written module by module. Learning that a release is a promise, and that data outlives every framework built on top of it.
2002 — 2015
Thirteen years of enterprise integration across four continents — middleware migrations, platform consolidations and the long education that the hard problem was never the code, it was the contracts between teams.
2015 — 2021
Lifting estates off the data-centre floor. Autonomous databases, IoT, pipelines and the discovery that architecture is mostly the art of making change cheap.
2021 — today
Model-led systems, knowledge graphs and retrieval built as governed enterprise platforms. The frontier moved again — and thirty years of fundamentals turned out to be exactly the right preparation.
The next crossing
Architecture reviews, cloud strategy, AI adoption, or simply an honest second opinion from someone who has seen this pattern before.