Behind private wealth sit confidential records, important relationships, critical systems and decisions that must remain sound over time. Exposure can arise not only from financial choices, but from how information is accessed, shared, retained and recovered.
Technology and AI decisions are getting larger, faster and harder to reverse. I provide an independent view when a commitment needs testing, and fractional CIO or CTO leadership when an organisation needs someone senior to take the technology agenda forward.
Independent advisoryFractional CIO / CTOSelected workThe thinking
Agron Fazliu · Independent technology adviser
The supplier has an answer. The internal team has one, often for good reasons. The consultancy has a programme. The security specialist wants more controls. They can all be partly right at the same time.
None of them necessarily approves the spend, explains the trade-offs to the board or runs the organisation that must live with the result. Most difficult technology decisions do not need another confident opinion. They need an independent read of what is actually true.
Start with the decision, not the product.
Before a major commitment, I examine the evidence, challenge the assumptions and set out the choices. This may be a technology decision review, AI authority assessment, supplier review or investor diligence. The engagement is bounded by the question, not by a delivery programme to sell.
Independent advisorySome organisations do not only need a recommendation; they need someone senior to own the technology priorities, challenge suppliers, direct a programme and report to leadership over time. I can take a defined, part-time or time-limited CIO or CTO mandate, with decision rights, cadence and handover agreed at the outset.
Fractional CIO and CTO leadershipParticularly relevant where private information, many external advisers, legacy systems or a growing investment portfolio make technology leadership consequential.
I work particularly in environments where confidentiality is part of the service, and where technology decisions have consequences well beyond the technology function.
An AI programme, cloud move, acquisition, supplier decision or architecture choice that will be expensive to undo.
Legacy systems, fragmented information, hidden dependencies or a transformation that has stalled without a clear explanation.
Questions of access, authority, confidentiality, accountability and how a system can be stopped or its actions reversed.
Security, supplier dependency and recovery that must be tested rather than accepted on paper.
This perspective comes from leading technology and risk functions, changing operating systems under constraint, and examining the human dependencies that project plans tend to miss. Much of the work takes place where discretion matters. The organisations are not named; the problem, reasoning and outcome are what count.
An ageing collection of systems had to be replaced before a failure made the decision. The whole estate moved on a fixed date, a paper archive became searchable text, and recovery was demonstrated in an exercise rather than assumed.
Once software can read confidential information and take action, it needs explicit limits on identity, access, authority and accountability, not simply a promising demonstration.
Communication patterns can show where knowledge and coordination actually sit, provided the analysis does not become a means of watching individuals.
AI gets more interesting when it leaves the demonstration and enters an organisation. The difficult questions are what it may see, what it may do, who is answerable when it is wrong, and whether anyone can reconstruct what happened.
The moment a system can act, it needs governance that takes its authority seriously. A useful AI system is not simply one that can do more. It is one whose limits someone has deliberately decided.
My work has moved through networks and infrastructure, security and risk, data, software platforms, product and executive technology leadership. The difficult decisions rarely stay inside one of those disciplines. Knowing how they connect changes the questions worth asking.
The background includes international technology operations, data and security leadership, CIO responsibility, research, and public-sector systems. Time across the technical and executive layers is useful only if it leads to clearer judgment today.
Former CIO; earlier responsibility for information security and IT risk, DevOps, data and product technology.
MSc, University College London · Executive education, University of Cambridge · Chevening Scholar
I write when a problem is useful beyond the project that produced it: about AI, data, organisations and the ways technology decisions go wrong.
AI adoption is as much a question of judgment and governance as capability.
Medium, 2024Why correlation is a poor substitute for explanation.
Medium, 2020How to use organisational data without forgetting the people behind it.
Medium, 2019The useful time for an independent view is before the decision becomes difficult to reverse.