In depth

Longer pieces on the technology leadership decisions we work on, drawn from our own searches.

AI at scale is gated by identity, not by models

Every AI agent needs an identity, a scope and an owner. Enterprise estates were built for human identities, and that is what is holding the work back.

In a portfolio company the security hire is often two jobs at once

The security role in a PE-backed business often has to be hands-on control building and board-facing judgement at once. That's where the brief goes wrong.

AI has changed what a software business is worth, and the buyer’s own team is why

AI has changed what software businesses are worth for buyers and sellers alike. The valuation now turns on what the customer could build without you.

A business that has outgrown its systems needs to know why before it briefs a CIO

A brief that reads "we've outgrown our systems" as a platform problem hires the wrong CIO, because why the systems fell behind is the mandate.

What a CTO appointment is hired to fix depends on why the role is open

A CTO appointment is shaped by why the role is open and where the owner sits in the hold, and one of the four mandates has nothing broken in it at all.

AI skills versus AI experience: how to hire for the work, not the words

A lot of the data and AI leadership work right now is sorting who has done the work from who has rebranded. The rebranding is not necessarily dishonest, and a recent AI title proves nothing either way. The test is the chronology behind the title and the layer of work underneath it.

Product and engineering tension decides where teams break or scale

Product and engineering carry different definitions of done and different owners, and nobody sits between them holding both. When one side wins every argument, architecture shapes the roadmap and product requests get tested against it. The hire that fixes it has to hold both definitions in the same role.

Who should AI report to? The reporting line is a commercial call

Where AI reports is the cleanest signal of who owns the commercial outcome from it. Across our placements it sits under one of two owners, and the choice follows what the business expects AI to do for revenue, margin or risk. The line carries the budget, the metrics and the escalation.

A CTO hire into an under-led function is the first of several hires

A PE-owned healthcare business brought in a CIO, then made six more senior appointments across product, data, applications, infrastructure, architecture and security. The brief could have been written as a standalone hire. The job was to build the team around them, and the next hires followed.

Your target-state operating model decides whether you need a chief architect

A chief architect earns its place when architectural decisions have to be made once and then held across more than one team, product or legal entity. Headcount is rarely the trigger. The number of decisions with no owner is, and until the target state is described the role is a name on an org chart.