Enterprise AI, from hype to working tools

Securitas is one of the largest security services organizations in the world, with more than 360,000 people globally. In North America alone it runs 580 offices and a field workforce of roughly 90,000, with service delivered by teams embedded at client sites around the clock.

I held a dual mandate: executive accountability for a $2.3B global client portfolio, and ownership of the firm's enterprise-wide AI and technology transformation agenda. That combination mattered. Because I also carried the portfolio's numbers, every technology decision had to earn its place in the operating results, not just in a slide deck.

The market around us was loud. Every vendor had an AI story, and every leader was being asked what their AI strategy was. My job was to cut through that and put tools in people's hands that actually changed how decisions were made and how work got done.

The test I applied to every AI idea was simple: does it change a decision, a forecast, or a workflow that someone in the business cares about this quarter?

Separate market noise from capabilities that would deliver a real return, at a time when the technology was changing month to month.

Serve a business that is distributed, field-heavy, and running 24/7, where any disruption to service delivery lands directly on a client.

Bring leaders and teams with very different levels of digital fluency to the same baseline, across multiple countries and time zones.

Keep governance tight for a portfolio that includes heavily regulated clients, including some of the largest US banks.

Build something that would last: a repeatable capability, not a string of disconnected pilots.

I set the vision and owned the enterprise AI adoption roadmap end to end: evaluating the market, selecting vendors, sponsoring deployments, designing the change program, and holding the function accountable for results. I was also personally hands-on with the tools, because I do not believe you can lead an AI program you have not used yourself.

Separate signal from noise

I built a practical filter for AI opportunities: start from a real operating problem, estimate the value, and only then look at technology. That kept us focused on pragmatic, ROI-positive use cases and away from demos that impressed but did not move the business. I evaluated generative AI platforms first-hand, including ChatGPT, Claude, Gemini, and Copilot, as part of vendor selection and adoption strategy.

Build the engine, not just the projects

I stood up a centralized Innovation and Transformation function from the ground up. It owned institutional playbooks, custom AI tooling, and the change-management programs that turned individual wins into standard practice. Having one accountable home for this work is what made it repeatable.

Put better data in front of decision-makers

I championed AI-powered analytics and automated reporting infrastructure on Power BI and Oracle BI, so leaders across the portfolio could see performance without waiting for someone to assemble it by hand. The goal was to democratize data-driven decision-making, not to centralize it.

Automate the work around the work

I evaluated and rolled out workflow automation with Zapier and n8n, PMO and portfolio-tracking platforms including Smartsheet and Airtable, and team AI enablement with Notion AI. These removed friction from the coordination work that quietly consumes senior leaders' time.

Bring people with us

I designed and rolled out AI-literacy initiatives across a globally distributed workforce. Adoption was treated as the real product: training, champions, clear use cases, and support, sequenced so that operations never felt the change as disruption.

Faster, better decisions

AI tooling and automated analytics improved decision-making, forecasting accuracy, and execution speed across a global organization.

Efficiency in the operating model

Automated efficiencies were embedded into the core operating model and contributed to expanding operating margin from 7.9% to 10%.

Zero operational disruption

A globally distributed workforce adopted new tools during rapid technological change with no disruption to client service.

A lasting capability

An Innovation and Transformation function with playbooks, tooling, and change programs that outlived any single project.

Start with the operating problem, not the model.

If you cannot name the decision or workflow that changes, you do not have a use case yet.

Adoption is the product.

A tool nobody uses has an ROI of zero. The change program deserves as much design attention as the technology.

Lead it by using it.

Hands-on fluency lets a leader ask better questions of vendors, teams, and the numbers.