Technology leadership. AI transformation. The people to deliver.

When technology needs leadership—not another recommendation.

A platform decision rarely stands alone. It affects the operating model, the team, the budget, and the organization’s ability to serve clients.

I assess the technology and the way work moves through the organization: where decisions stall, where teams depend on one another, which systems constrain growth, and which capabilities are missing. From there, I establish priorities and lead the changes required to execute them.

That work can include:

• Technology strategy, platform assessment, and build-versus-buy decisions.
• Modernization of legacy applications toward distributed, cloud-ready architectures.
• Cross-functional planning, clear ownership, and more predictable delivery.
• Team assessment, technical leadership, and additional delivery capacity.
• Leadership of an existing technology function or a defined transition and rebuilding mandate.

At Zelis Healthcare, a healthcare payments and technology company, my mandate combined modernization of a legacy claims-pricing platform with AI transformation and additional delivery resources. Earlier, as Fractional CTO through Emerging Technologies, I helped IPG modernize its platforms and improve its delivery operating model.

Change how the team works—not just which tools it uses.

Individual AI tools can accelerate isolated tasks without improving the flow of work across product, analysis, engineering, and QA.

I embed AI into the team’s existing delivery process: planning, refinement, implementation, testing, review, and retrospectives. The aim is to improve delivery while preserving accountability, quality, and the domain knowledge the work depends on.

Build security and IP protection into the adoption path.

Security review and intellectual-property concerns are part of AI adoption, not a final approval gate. I work with technology and security leaders to define approved models and hosting, data and code boundaries, retention expectations, identity and least-privilege access, audit needs, and human approval points before expanding automation.

Where MCP is appropriate, it can provide a governed interface between agents and approved tools or systems. MCP is not a security control by itself: authorization, scoped permissions, validation, logging, and ownership still need to be designed for each connection.

The team should not have to maintain a second documentation program to train AI. Context should emerge from useful work already happening—and return value through less repeated explanation, fewer recurring mistakes, and faster onboarding.

At Zelis, this approach demonstrated 4× engineering throughput, with 10× improvements on specific delivery tasks. Those are observed results from that engagement, not a blanket forecast for another organization.

I start with the decisions the organization needs to make, identify authoritative sources, map the relevant relationships, and validate them with subject-matter experts. Access controls, ownership, review, and maintenance are part of the design.

Delivery agents need business context that source code cannot explain. Leadership needs an accurate view of systems and delivery constraints. Team context and enterprise knowledge inform one another without becoming one undifferentiated repository.

You do not need to buy a transformation package before addressing a specific problem.

Where the direction is unclear, a 3–4 week AI Strategy Diagnostic can establish readiness, priorities, build / buy / wait decisions, and a practical 90-day sequence. No rate card—scope follows the mandate.