The difficult part of AI is not accessing a model.

It is helping an organization become capable.

Light Air is an independent AI enablement and implementation practice. I help organizations identify useful work, introduce purpose-fit models and narrow agents, and teach the people already doing the job how to supervise and improve them.

The goal is not to make a business dependent on Light Air—or on a commercial AI laboratory. The goal is to leave the organization more capable than we found it.

A background in operations and enablement

My work was shaped by more than a decade in operational, customer-facing, and leadership roles at Pine Environmental. I worked across sales operations, process performance, analytics, business-unit support, and regional management.

Much of that work involved understanding how information moved through an organization, improving the systems around it, and helping people learn to use ERPs, CRMs, and new processes effectively. I became comfortable standing in front of a room, making complicated systems understandable, and helping people move from uncertainty to practical confidence.

That experience is the foundation of Light Air. AI adoption is not primarily a software problem. It is an organizational learning problem.

Begin with the work

Light Air does not begin with the newest model or the loudest promise. We begin by asking:

Where is time being lost?
What knowledge is difficult to find?
What routine work keeps pulling people away from customers, decisions, and one another?

Then we choose one useful job, build carefully, establish clear human authority, document the system, and teach the people who will live with it. When local AI fits the work, the organization should be able to own the infrastructure, protect its operating knowledge, and develop capability that improves over time.

Purpose-fit AI, built with care

I believe AI should strengthen people rather than make them disposable. A useful system should be understandable, appropriately bounded, and honest about what it can and cannot do.

It should reduce administrative drag while preserving human judgment. It should give an organization greater control over its knowledge—not create another permanent dependency.

Light Air uses established models and open tools, and brings in specialized technical support when a project genuinely requires it. We do not pretend every problem needs AI, and we do not promise transformation overnight.

Let’s begin with the work

If there is a process your people are tired of carrying, knowledge they cannot reliably reach, or a routine task that keeps stealing attention from more valuable work, tell me about it.

It's free to talk →