Blog

Notes from programmes that had to justify themselves.

Opinions formed inside regulated estates, real budgets, and production load. Where I disagree with the industry, I have tried to say so plainly.

Data & Digital Platforms3 min read

Latency is a business metric, not an engineering one

Reporting that arrives the next working day cannot inform a decision made today. Cutting a data platform from ~12+ hours to ~15 minutes did not just make a dashboard faster. It changed who was allowed to act, and when.

Read

Enterprise Engineering2 min read

The licence is a tax you stopped noticing

High-licence COTS platforms are usually justified as buying capability. Often they are renting something the enterprise already has the talent to build, and the roadmap is what you are really paying for.

Read

Data & Digital Platforms2 min read

Open source is an architecture decision, not a cost decision

The large data lakes I have built stand on open-source products because they compose better. The savings are a consequence of that choice, not the reason for it, and treating them as the reason leads you to build the wrong thing.

Read

AI & Autonomous Networks2 min read

What Level 3 autonomy actually demands of your organisation

The hard part of an autonomous network is not the models. It is agreeing, in advance and in writing, what the system is allowed to do without asking, and building the organisation that can live with that answer.

Read

Business Transformation2 min read

Transformation must ultimately connect to measurable business value

Every transformation programme should carry a number it is accountable to. Not a proxy, not an activity metric. A figure a finance partner will put their name against.

Read

Regulatory & Compliance2 min read

Compliance is not the reason nothing ships

In regulated, public-company environments, compliance becomes the accepted explanation for slow delivery. It is usually the alibi rather than the cause, and treating it as the cause guarantees it becomes one.

Read

Telecom / AI3 min read

Why Autonomous Networks Are an Operating-Model Problem, Not an AI Problem

Most autonomy programmes are funded as model programmes and stall as organisational ones. The constraint is rarely what the system can decide. It is whether the operator has redesigned who owns the decision once the system makes it.

Read

CEO Strategy2 min read

The Telecom CTO Is Becoming a Business Strategist

The technology leader's job used to be keeping the network running and the programmes on budget. In a business where the network is the product, that job is now inseparable from strategy, capital allocation and growth.

Read

AI2 min read

The Difference Between an AI Demo and an AI Operating Model

A demo proves that a model can do something once. An operating model proves that an organisation can rely on it every day. Much enterprise AI spending still buys the first and hopes for the second.

Read

Data2 min read

Why Data Latency Is a Revenue and Customer-Experience Problem

Late data is usually filed under engineering efficiency. Its real cost shows up elsewhere: in problems customers notice before the operator does, in capacity decisions made blind, and in commercial opportunities that expire before anyone sees them.

Read

M&A2 min read

The Technology Due-Diligence Questions Every Telecom Acquisition Should Ask

Technology due diligence is often treated as a checklist that confirms the deal. Its real job is to test the investment thesis: whether the synergies are achievable, what integration will actually cost, and which risks the price has not accounted for.

Read

Capital / Technology Strategy2 min read

When Should a Telecom Operator Build Instead of Buy?

Build versus buy is usually argued on cost. It should be argued on control: which capabilities differentiate the business, how fast they need to change, and whether the operator is prepared to own them for the long term.

Read

Organization / Leadership2 min read

Why ~450+ Engineers Don't Need More Strategy Decks—They Need a Better Operating Model

Large engineering organisations rarely fail for lack of strategy. They fail because the strategy never reaches the way work is actually organised, funded and decided. At that scale, the operating model is the strategy.

Read

Speaking

These arguments travel well to a stage or a podcast.

If a theme here is useful to your audience, I am glad to take it further in person.

Start a conversation