Evidence before claims.

Enterprise Data and AI operating system

Situation

A diversified manufacturer faced rapidly increasing Data and AI demand across mills and corporate functions. Responsible scale required common direction, business ownership, trusted data, governance, platform foundations and adoption.

Why the context mattered

Manufacturing transformation must respect operational continuity, safety, cybersecurity, legacy environments, site-level realities and the expertise of people closest to production. The challenge was not to impose a fashionable AI agenda. It was to create an enterprise system that could support local value, common standards and responsible learning at the same time.

What I built

• Enterprise strategy and multi-year roadmap
• Governed investment and use-case portfolio
• AI Steering Committee and execution subcommittees
• Agent Center of Practice and secure experimentation
• AI risk model, minimum guardrails and lifecycle
• Azure and Microsoft Fabric Data Product Foundation
• Governed OEE and unplanned-downtime data products
• AI Academy, practitioner community and adoption streams

Executive judgment applied

I connected corporate ambition to mill-level ownership, experimentation to minimum controls, platform investment to reusable data products and enthusiasm to a governed portfolio. The result was not a collection of pilots, but the beginnings of a repeatable enterprise capability.

Evidence status: Strategy, governance, platform and adoption capabilities are established and operating. Specific enterprise benefits remain subject to progressive operational validation.

Lesson

AI scale is not primarily a model-selection problem. It is an operating-model problem involving ownership, trusted data, decision rights, controls, adoption and evidence.

AI for global maritime operations

Situation

A global maritime organization operating in an asset-intensive, safety-sensitive environment needed to modernize its digital foundations and improve operational decisions.

Beyond the AI use cases

As VP IT & Digital, my accountability extended across applications, infrastructure, vessel and shore connectivity, cybersecurity, architecture, data and digital delivery. That breadth mattered: an optimization model is valuable only when data arrives reliably, operational users trust the recommendation, security is preserved and the surrounding technology can support the decision at sea.

What we delivered

• Maritime operational control tower
• Improved vessel ETA prediction
• Digital twins for operational decision support
• Speed and route optimization
• Modernized warehouse-to-lakehouse environment
• Approximately 1,500 data packages and views
• Roughly 2,000 IoT signals per ship every five seconds
• Data-sharing and cybersecurity standards

Executive judgment applied

We treated domain experts, data engineers, technology teams and operational leaders as one delivery system. We paired ambitious innovation with the reliability, funding, security and adoption disciplines required in a global industrial environment.

Outcome note: The work enabled 15–20% fuel reduction through route and speed optimization. This should not be represented as universally realized savings for every vessel or voyage without applicable operational evidence.

Lesson

Industrial AI creates value when domain knowledge, data engineering, workflows, technology reliability and business ownership are designed together.

Turning fragmented information into commercial capability

Led technology across a large automotive group, built Canada’s first comprehensive used-car data aggregation platform and developed products that standardized third-party data for valuation, inventory, content and market decisions.

The operating complexity

Automotive connected consumer traffic, dealer operations, inventory, valuation, third-party feeds, digital media, content and commercial partnerships. The work required platform thinking, data normalization and customer experience—but also revenue logic, service reliability and an understanding of how technology changes front-line decisions.

What this experience added

It taught me to view data as a product with consumers, quality expectations and economic purpose. It also reinforced that digital scale magnifies both strengths and weaknesses: unclear definitions, weak integration and poor ownership become more costly as audiences grow.

Lesson

Data becomes valuable when it is reliable, understandable and delivered inside a decision or workflow. Aggregation alone is not enough.

Creating something from nothing

Ensigna

Built a distributed IT-services model using a network of more than 100 students. The company grew to 12 full-time employees and approximately $800,000 in annual revenue before being sold in 2000.

Eurekium Innovations

Built an enterprise mobile middleware company that grew to 53 employees, four European offices and approximately $6 million in investment. Clients and ecosystems included AXA, SOS Médecins, Orange/France Telecom, Capgemini and Deutsche Telekom. The company was sold in 2004.

What founders learn early

A founder cannot outsource the connections between proposition, product, sales, delivery, financing, people and customer trust. I learned to make decisions with incomplete information, to distinguish interest from demand and to understand that execution problems eventually become commercial problems.

How clients benefit now

I bring that ownership mindset to enterprise work. I look beyond whether an idea is technically possible to whether someone needs it, will fund it, can operate it, will adopt it and can prove the outcome.

Lesson

Entrepreneurship connects vision to cash flow, product ambition to market reality and innovation to delivery discipline. It also teaches humility quickly.