I'm a fractional Chief AI Officer with 20+ years as a CTO and VP of Engineering. I help small and mid-sized companies turn AI from a boardroom conversation into automated operations, autonomous agents, and measurable margin, without betting the company on hype.
Products and platforms delivered for
This is not a technology cycle you can sit out and adopt later at a discount. The companies automating today are building an operating advantage that gets harder to catch every quarter.
Competitors running AI-driven operations are removing entire categories of manual work. When they can serve the same customer at a fraction of your cost per transaction, your margin stops being a strategy problem and becomes a survival one.
Quoting, onboarding, support, reporting, compliance: every process an AI agent handles runs in minutes instead of days. Customers notice. Deals go to whoever answers first with the right answer.
Automation encodes your processes, your data, and your institutional knowledge into systems that keep improving. Start two years late and you are not two years behind. You are behind by everything they learned in those two years.
The question is no longer whether AI reshapes your industry. It is whether you are the one doing the reshaping.
Find your first moveStrategic enough for the board. Technical enough to earn the engineering team's trust. Practical enough to ship in weeks, not quarters.
AI leadership on your executive team without a full-time hire. I own the AI agenda: where it goes, what it costs, what it returns, and who builds it.
We take the most inefficient operations in any department and automate them end to end with production-grade systems, not brittle scripts and not a demo that dies after the pilot.
Autonomous agents that make decisions inside real workflows and SOPs (sales, support, operations, finance), with the guardrails, evaluations, and audit trail a business can actually stand behind.
The decisions that cannot be postponed: scale, cloud, security, platform redesign, legacy migration. I have led these transformations for enterprise platforms and I lead them here.
A rigorous, independent analysis of the AI projects on your table, before you commit budget. What is genuinely viable, what the real cost is, and what it returns.
Your buyers increasingly ask ChatGPT, Claude, Perplexity and Google's AI Overviews instead of scrolling a results page. If those models cannot find you, parse you, or trust you, you are invisible in the fastest-growing channel in your market.
Content written for two audiences at once: the person making the decision, and the model summarising you to them. Built from your real expertise and structured so machines quote you accurately instead of guessing.
Typical engagements run one to three days a month. No long lock-in.
Book a callYou do not need a transformation programme to start. You need one process fixed properly, proof that it works, and a plan for the next five.
Thirty minutes. You describe how your operation actually runs. I tell you where AI pays back fastest and where it is a waste of money. No slides required from either of us.
Two weeks. I map your processes, rank them by payback period, check technical and data feasibility, and hand you a costed plan. It stands on its own whether or not you continue with me.
I lead delivery, embedded with your team or with mine, starting with the highest-return process. Then we repeat, and your people take ownership as we go.
AI systems in production for real companies, plus the products I have built and shipped myself. Every one had real users and real revenue at stake.
CoachSafely trained youth coaches the way most organisations still do: a library of course videos, watched once, and half-remembered by the time it actually mattered on the field. If a coach faced a situation the video did not cover, there was nobody to ask.
We built an AI agent that learned the entire course and the underlying health and safety protocols, then delivered it through a mobile app connected to our own model. The agent holds conversation history, so it knows what each coach has already covered and adapts to them over time.
The result is not a chatbot bolted onto a course. Every coach now has a private teacher who is available at any hour, answers questions about the specific situation in front of them, and stays grounded in the organisation's own protocols rather than guessing.
Visit CoachSafelyEvery department had its own system, and every system had its own learning curve. Answering a routine question meant knowing which tool held the data, having access to it, and knowing how to get it out. Most people did neither, so requests queued behind the few specialists who did.
We built an agent connected directly to the company database, its documents, and its recorded customer conversations. It answers questions in plain language across all three, and it does not stop at answering. It executes actions back in the underlying systems.
Nobody has to learn the software any more. An employee describes what they need and the agent does the rest, which removes software training as a bottleneck and multiplies what one person can get through in a day.
I founded K8Studio and led product and technology from idea to a cross-platform desktop product: multi-cluster operations, CloudMaps, security, observability, Helm, ArgoCD, and an AI copilot in one local-first workspace.
I created UxxU to connect C4 systems, containers, components, and dependencies in one collaborative workspace. Architecture becomes navigable, measurable, embeddable, and usable by AI through MCP and code-to-C4 workflows.
As VP of Software Development I led five cross-functional teams (100+ people) replacing a legacy Flex stack with Angular, React and HTML5, and migrating the backend to microservices. Platforms used by MGM Resorts, Royal Caribbean and Carnival.
Advisor and Chief Solution Architect: multi-tenant platform architecture, scalability, performance and security, data aggregation, and hands-on technical mentoring across Node, React, Angular, TypeScript and Python.
Why software architecture needs a living connection to the code that AI is changing.
Read article
The hidden security and operational trade-offs behind convenient infrastructure tooling.
Read articleWhat it takes to turn thousands of workloads into a product engineers understand in seconds.
Read articleA practical approach to secure private cluster access without creating operational drag.
Read article
I'm a technology executive, architect and product founder with 20+ years turning complex systems into business momentum, and for the last few years, turning AI into operations that actually run.
I've led engineering organisations of 100+ people across the US, Europe, Australia and New Zealand, delivering production platforms for global brands including MGM Resorts, Royal Caribbean and Carnival.
I founded UxxU and K8Studio, so I know what it costs to build and ship a product, not just advise on one. I work comfortably from board strategy down to production architecture.
Far less than a full-time executive and a fraction of a large consultancy. Most engagements run one to three days a month on a fixed monthly retainer, so the cost is predictable from day one. The two-week diagnostic can be bought on its own if you want proof before committing to anything ongoing.
The diagnostic takes two weeks and gives you a costed, ranked plan. First automation in production is typically four to eight weeks after that, depending on how clean the integrations are. If someone promises you production AI in a week, they are selling you a demo.
Almost everyone's data is a mess, and waiting until it is clean is how companies lose three years. We start with processes where the data you already have is good enough, get value on the board, and use that momentum to fund the cleanup that genuinely matters.
Small and mid-sized companies are exactly where this works best. You have less bureaucracy, shorter decision chains, and processes that a well-built agent can absorb entirely. Large enterprises spend a year in committee to do what you can ship in a month.
I sign your NDA before we talk specifics. Architecture decisions account for where your data lives, which models can see it, and what your regulatory position requires, including private, self-hosted or EU-resident models where that is the right call. Twenty years of building for enterprise clients means I treat this as a design constraint, not an afterthought.
Often it makes more sense. A good CTO is already at capacity running the product and the team, and AI transformation is a second full-time job. I work alongside your CTO as a specialist, not over them, and part of the mandate is making sure your team owns what we build.
Bring me your operation, your roadmap, or the problem everyone is avoiding. Pick a time that suits you. The calendar below is live.
No pitch deck. No procurement dance. If it is not a fit, I will say so on the call and point you somewhere better.