An investment remains a responsibility after the vote

A board's technology discussion becomes more useful when it concerns the business the company is choosing to own. That may include a common product, a specialist client version, an external component or an experimental capability whose value remains uncertain. I would help directors ask what continuing responsibility sits behind each choice. The aim is not to make the board an architecture committee for every programme. It is to give collective oversight a clearer account of the investment, the alternatives and the conditions on which management's recommendation depends.

My former directorship at Shenlume Instruments ran from July 2018 to June 2022, alongside membership of its technology committee. That service was separate from my executive work at Lanshuren. It taught me to contribute questions without assuming the operating authority I hold in my own company. A director needs to understand enough to challenge a proposal and recognise its uncertainty, while leaving management responsible for implementation. I do not turn that experience into a claim to chair an audit committee or to assess every financial and legal issue without other expertise.

Product consolidation is one useful example of the questions I would bring. Moving sixteen runnable branches to six can look like a straightforward efficiency result. The more consequential account explains why five specialist lines remain, who owns them and whether their continuing support appears in the investment case. Directors do not need to choose the configuration mechanism. They do need to know whether the proposed benefit depends on obligations that management has omitted or on client changes that have not been agreed. A smaller number of branches is evidence about architecture, not by itself a complete statement of commercial value.

An AI evaluation creates a different oversight question. Our fixed-set exercise produced a lower measured inference charge per accepted output. I supported a limited trial because the result concerned a defined task and familiar population. The board-level issue is the scale of the decision management wants that evidence to support. If a narrow experiment is used to justify a broad investment, directors should ask what bridges the gap. That includes the intended population, the acceptance criteria, the excluded operating costs and the evidence that would cause management to narrow or stop the proposed use. A confident demonstration should not relieve the company of explaining those matters.

A purchased component illustrates why reduced engineering allocation is not necessarily a financial saving. In the scheduling decision, engineers remained employed and some moved to domain workflows. Licence charges and support obligations continued. I would want the investment account to explain the value expected from that redirection and the responsibility retained for the supplier relationship. The demonstrated export route addressed one exit uncertainty, while other switching costs remained. A board can examine the credibility of that account without deciding that internal build or purchase is inherently more responsible.

The same distinction applies to reuse. Engineering may see an attractive component from a previous engagement, but technical usefulness does not determine permission. In our fourteen-component review, authorised colleagues established which nine could be used for the stated product purpose and which five remained outside it. Seven of the nine were then selected for particular programmes. Approval, availability and adoption were different states. I would help directors ask whether management preserves those distinctions when it describes the product's assets, without asking the board to interpret each agreement itself.

For a new appointment, I would begin with the company's economics, customers and material obligations rather than assume that these examples define its agenda. Software sold as a product, embedded in equipment and used to provide a service creates different relationships and risks. Induction should include finance, commercial operations and the company's governance arrangements, with room to acknowledge what I do not yet understand. I would seek feedback from the chair on whether my questions improve collective judgement and whether I give sufficient space to other directors' perspectives.

I have no current directorship and would plan twenty-four annual days only after examining the actual calendar, employer consent and potential relationships. Previous service does not establish independence for a different company. Relevant client, supplier, employer and advisory connections would require assessment with the company secretary and appropriate advisers, and the account would need to remain current. I would join a suitable technology or strategy committee where my contribution fits the board's needs. The purpose would be sustained oversight of consequential choices, supported by preparation and honest limits, rather than an additional executive role under a different title.