Perspectives on
enterprise AI.

A board asks "are we ready for AI?" and gets back a feeling, a vendor's opinion, or a forty-slide deck. The Prime Diagnostic Score turns that question into a single defensible number. This piece shows the working — five PRIME pillars, each out of 20, summed to 100, mapped to five readiness bands — and what each band means for your next move.

Most firms name themselves after a founder, a place, or a promise. We named ours after a number — because 73 says something a surname never could. The meaning of the name, why we built for the APAC mid-market, why now, and the principles we don't bend on.

Most mid-market AI failures are data failures wearing an AI costume. The demo runs on clean sample data; production runs on yours. This piece sets out what "data ready" actually means — accessible, accurate, governed, connected — and why the foundation has to be assessed before a build is funded against it.

Directors don't need to be technologists to govern AI well — they need to ask the right questions. Five plain questions that surface risk and discipline before the cheque is signed: the problem and its metric, the exposure, the owner, the failure plan, and the cost of doing nothing.

Most vendor selection happens watching a demo — and the demo is the one thing the vendor fully controls. A practical checklist for non-technical leaders: what happens to your data, what the system does on your data rather than theirs, what it costs to leave, and why the evaluator shouldn't be paid by the answer.

APAC mid-market firms sit in a specific gap: enterprise-grade ambition on SME-grade foundations. This piece maps where the gap shows up — data, governance, talent depth — and how the regional texture of Australia, Singapore, and Hong Kong shapes what a single AI policy has to satisfy.

A readiness assessment looks like a cost before the real work begins. It is usually the highest-return dollar in an AI programme — because it stops the wrong problem, the premature commitment, and the mid-build surprise before any of them are funded.

The build-versus-buy question in AI is almost always framed wrong. The real question is which layers you build versus purchase — and the answer differs by organisation size, data maturity, and competitive context. This piece provides a decision framework for APAC mid-market firms choosing between SaaS tools, no-code automation, and custom development.

Most boards are approving AI budgets without the frameworks to govern them. This piece covers the minimum governance architecture a board should have in place before sanctioning significant AI spend — including committee design, regulatory obligations under Singapore's PDPA, Australia's Privacy Act, and Hong Kong's PCPD guidance, and the metrics that separate meaningful progress from theatre.

The quarterly cost of delayed AI adoption compounds faster than most boards appreciate. This piece quantifies the competitive lag for mid-market firms in Australia, Singapore, and Hong Kong — and argues that the cost of starting 12 months later is not linear.

The AIRA methodology — five PRIME pillars, one composite Prime Diagnostic Score, one board-ready report. This piece explains how the framework was designed, what each pillar measures, and why the composite scoring approach gives boards a clearer picture than pillar-by-pillar assessment alone.

Three preventable failure modes account for the majority of mid-market AI pilots that fail to reach production. All three are identifiable before a project starts — and all three are addressed in the Prime Diagnostic™ before a single dollar of implementation spend is committed.