Use AI effectively
On the jobs it is genuinely good at, and plain software for the rest. The judgment to tell which is which.
A credential proves someone studied AI. It does not prove they can use it, explain it, or keep it from going wrong in the real world. I am a working software engineer with 10+ years building reliable systems in regulated, high-stakes domains, and I learned AI itself the hands-on way: months of immersive building, testing, and breaking it, well past day-to-day use. So I teach what holds up in production, not what sounds good on a slide.
What really sets me apart is the lens I bring. For more than a decade I have studied geometry and patterns on my own, and I apply that directly to AI, because AI represents meaning as geometry, as shapes and directions rather than a list of facts. Seeing it that way is not a gimmick; it is how the thing actually works. I rarely see anyone teach it like this, and it is exactly why my explanations click instead of staying abstract. People do not leave with theory. They leave finally understanding how AI works.
On the jobs it is genuinely good at, and plain software for the rest. The judgment to tell which is which.
Tokens are money, and most teams burn them paying a thinking engine to do work a simple script should do. I teach where the cost hides and how to cut it.
Over-automation, unchecked hallucinations, complexity that only adds cost, and trust quietly eroding until no one believes the output.
I tell you what AI cannot do as plainly as what it can. No magic, no fear, no promises it can never be wrong.
Two episodes from the series, one for range, one for the business side. The full series is on the series page.
Six modules that build from the ground up, each anchored in concrete, everyday examples. No coding background required; technical staff get the depth they rarely hear stated out loud.
AI as a pattern and prediction engine, not a mind or a search box; the myths cleared; the four behaviors that explain everything else.
Why AI is confidently wrong, how to catch it, and a repeatable trust process: ground it, make it cite, check every claim against the source.
What the human brings, practical prompting, and writing a standing directive (role, values, boundaries) that persists across sessions.
The model is the engine; match it to the job by size, cost, and privacy; meaning as geometry; what an agent is and the governance it needs.
What a token is and how billing works; the costly anti-pattern of automating deterministic work with a paid model; the doing-versus-thinking rule.
Use AI to simulate a plan before you spend a dollar, then build it lean so it stays cheap and clear.
The video modules in order, with the takeaways as a reference. Free to watch on the series page.
Each module as a guided workshop with Q&A and hands-on exercises, for a team or a group.
Tailored to your organization's tools, policies, and real use cases.
The series was produced on a custom browser-based studio I built using AI-assisted development: script cueing, on-camera coaching, and timeline editing. It is itself a working demonstration of the AI-assisted engineering taught in the later modules, the teaching comes from practice, not slides.
If your organization wants its people to use AI well, accurately, and affordably, that is what I do. Let us talk about a workshop, a cohort, or a tailored curriculum.