Procurement · Australia
AI-ready procurement: buying technology your business can actually grow into.
Most technology being bought in Australia right now will still be in service in 2030. Very little of it is being chosen with the AI workloads of the next five years in mind.
AI-ready procurement is the discipline of closing that gap: choosing devices and infrastructure that can run the tools your business is about to adopt, at a lifecycle cost you can defend to a board, a bank or a client.
What AI-ready actually means
It is not a badge on a product line. It is a set of practical questions asked before the purchase order: can these devices run a local model if privacy later demands it? Is there enough memory for the tools on your roadmap, not just today's? Can the platform be patched and supported for the length of its depreciation? And can you leave the vendor without rebuilding everything?
Those questions cost nothing at evaluation and a great deal after delivery.
Why refreshes go wrong
The usual failure is buying to a spreadsheet. The cheapest compliant option wins, the specification is written around last generation's workloads, and eighteen months later the business finds the devices cannot run the assistant it just licensed.
The second failure is the opposite: paying for headroom nobody will use because a vendor was persuasive about future-proofing. Good procurement sits between those two, grounded in what your team will actually run.
How an assessment works
I start with the workloads: what your team does now, what tools are being trialled, and what is likely over the next two to three years. Then the constraints: budget cycle, warranty expectations, security and data residency requirements, and who will support the fleet.
From there you get a documented specification, an independent comparison of the options, and a lifecycle cost view across the refresh period. If you already hold quotes, I review them and tell you what is over- or under-specified.
What this looks like in practice
Device fleet refresh
Laptops and desktops specified for current and near-term AI workloads, with a defensible replacement plan.
Infrastructure sizing
Server, storage and network requirements derived from real workloads rather than a vendor bundle.
Vendor evaluation
An independent scorecard across cost, support, lock-in and product roadmap before you commit.
Quote review
An existing proposal checked for over-specification, gaps and the true cost across its service life.
Questions people ask
What is AI-ready hardware?+
Hardware that can run the AI tools your business is likely to adopt over its service life, including local inference where privacy requires it, while staying supported, patched and scalable for the length of its depreciation.
Do you sell hardware?+
The engagement covers specification, comparison and quote review rather than resale. The purchase is made by you, through your preferred supplier or one I help you evaluate.
When should procurement start?+
Before the specification is written. Once a spec exists, the decisions that matter — architecture, memory, lifecycle and lock-in — are already locked in.
Can you review quotes we have already received?+
Yes. A second opinion on an existing proposal is one of the fastest ways to find savings or to spot a device fleet that won't run what is coming.
Do you cover devices and infrastructure?+
Both. End user devices, and data centre and server infrastructure including storage and networking, sized to the workloads you actually expect rather than assumed headroom.