Our ideas from the forefront
of the AI revolution
We publish what we are learning as we build. Sharp thinking on AI transformation, agent governance, context engineering and where enterprise AI actually pays back, straight from the work, ahead of the curve.
New Zealand's AI Data Centre Debate: The Middle Path Both Camps Are Skipping
The Southland data centre fight has split into a climate camp and a growth camp, and both are arguing about the same asset. Why renewable electricity is simultaneously the climate answer and the commercial answer, and how New Zealand could lead on climate-safe AI infrastructure.
James Oldham10 min read4 Aug 2026Want to Run AI Agents Locally? The One Number That Decides Everything
A plain guide to running AI models on hardware you own. Why VRAM, not chip speed, is the constraint that decides everything, what quantisation buys you, the three real hardware tiers, and why the smart 2026 setup is hybrid.
James Oldham12 min read15 Jul 2026You Are Paying Frontier Prices for Commodity Work
Most AI budgets blow out for one reason: every request goes to the most expensive model in the building because nobody decided otherwise. How semantic model routing and conversation-level cost tracking cut the same workload by 20 to 40 percent.
James Oldham9 min read15 Jul 2026Your AI Vendor Is Building an Asset. It Is Made of Your Business.
Every prompt your team types teaches someone's system. The question is whose. Why AI sovereignty means owning the knowledge layer, keeping data in your region, and renting the model without renting out the memory.
James Oldham9 min read15 Jul 2026Why 95 Percent of AI Pilots Die, and What Saves the Other Five
MIT found 95 percent of enterprise GenAI pilots deliver zero measurable return. Gartner expects 60 percent of AI projects to be abandoned on data that is not AI-ready. The failures share one cause, and it is not the model.
James Oldham10 min read15 Jul 2026Agent Sprawl Is the New Shadow IT
Every team is quietly running AI agents. Nobody can list them. Why unregistered agents are a bigger problem than unregistered SaaS ever was, and why the register has to come before the governance.
James Oldham8 min read15 Jul 2026Best Enterprise AI Development Agencies in Australia and New Zealand (2026)
An honest comparison of the AI development agencies and platforms available to Australian and New Zealand businesses in 2026. Who each one is best for, what they cost, and the questions that separate builders from deck factories.
James Oldham11 min read15 Jul 2026What an Agent Management Platform Actually Is (and Is Not)
A plain definition of the agent management platform category: one control plane to register, monitor, govern and optimise every AI agent a business runs. What it is not: observability, a gateway, or another agent builder.
James Oldham9 min read15 Jul 2026Choosing an AI Agency in Wellington: A 2026 Buyer's Guide
How to choose an AI agency in Wellington in 2026. What good looks like, what to walk away from, and what an honest engagement should ship in 90 days.
James Oldham10 min read25 Jun 2026AI for Accounting Firms in New Zealand: The 2026 Operator's Guide
What AI for accounting firms in New Zealand actually looks like in production. Where it pays off, what to build versus buy, and how to vet a vendor in 2026.
James Oldham10 min read23 Jun 2026Fractional Chief AI Officer in New Zealand: The 2026 Playbook for Mid-Market Owners
A 2026 playbook on hiring a fractional Chief AI Officer in New Zealand. What the role does, what it costs in ANZ, and how to vet one before you sign.
James Oldham11 min read13 Jun 2026The Claude Code First Organisation: How to Build Your Company Around AI in 2026
Most organisations bolt AI onto broken workflows and wonder why it fails. The companies winning with AI in 2026 are doing something different. They are restructuring how their teams think, communicate, and operate so that AI models like Claude can actually reason across the business. This is what a Claude Code first organisation looks like in practice.
James Oldham14 min read9 Mar 2026Why Enterprise AI Fails Without Structured Context
Most AI projects fail not because the models lack capability but because the organisation's knowledge is fragmented, undocumented, and impossible for AI to traverse. The context problem is the real bottleneck of enterprise AI adoption, and almost nobody is solving it.
James Oldham10 min read4 Mar 2026AI Agent vs Off-the-Shelf Solutions: How to Decide (2026 Guide)
Off-the-shelf AI agents suit linear, low-volume workflows. Custom wins the moment an agent needs live access to your data, deep branching logic, or compliance-grade control. The decision framework we use with clients, with voice AI as the sharpest example.
James Oldham11 min read9 Mar 2026From Karpathy's LLM Wiki to the Company Knowledge Layer: Why Every Organisation Needs Its Own Knowledge Graph
Andrej Karpathy proposed an LLM wiki for personal context. The same idea applied at the organisational level is the missing piece behind every failing enterprise AI deployment. This is how a unified company knowledgebase plus a custom knowledge graph becomes the foundation of a company's AI transformation, and why the future is custom company LLMs grounded in your real data.
James Oldham13 min read22 May 2026Enterprise AI Consultancies for Retail in New Zealand (2026 Guide)
How to evaluate enterprise AI consultancies for retail in New Zealand in 2026. What to look for, what to avoid, and how the right AI partner can ship working AI inside your retail business in 90 days.
James Oldham11 min read26 May 2026AI Solutions in Christchurch: A 2026 Guide for Mid-Market Businesses
AI solutions for Christchurch businesses in 2026: what mid-market manufacturers, agritech and professional services firms are actually deploying, what real implementations cost, and the five-point checklist for choosing a provider.
James Oldham9 min read27 May 2026AI Solutions in Auckland: A 2026 Guide for Mid-Market Businesses
Practical guide to AI solutions in Auckland in 2026. What mid-market and enterprise teams are actually deploying, what to avoid, and how to ship working AI inside your Auckland business in 90 days.
James Oldham9 min read5 Jul 2026Choosing an AI Agency in Auckland: A 2026 Buyer's Guide
How to choose an AI agency in Auckland in 2026. What good looks like, what to walk away from, and what a serious AI implementation should produce in the first 90 days inside your Auckland business.
James Oldham10 min read28 May 2026AI Voice Agent for Recruitment: What Actually Works in 2026
What an AI voice agent does in a recruitment workflow, what separates production systems from demos, what it costs, and how agencies deploy it. Based on a live agent that has run 10,000+ candidate calls.
James Oldham10 min read29 May 2026Best Voice AI Platforms in ANZ (2026 Comparison)
Honest comparison of the best voice AI platforms available to ANZ businesses in 2026. What each one is good for, what to avoid, and when to build custom on Retell or ElevenLabs versus using an off-the-shelf platform.
James Oldham12 min read30 May 2026