I work at the intersection of AI, people, workflows, and real business problems.

My background is a little unusual. I’ve spent years in real estate, training, coaching, public speaking, curriculum development, small business, and customer-facing work. More recently, I’ve been building and testing AI-enabled workflows and products hands-on.

What interests me most about AI is not just what the model can do.

It’s what happens when you put AI into an actual business process.

Does it understand the situation? Does it know what it should do next? Does it know when it should stop? Does it know when a human needs to step in? Does the person using it actually trust it enough to adopt it?

Those are the kinds of problems I like working on.

• AI workflow design
• Human-in-the-loop systems
• AI evaluation and failure testing
• Prompt and instruction design
• Customer and employee AI adoption
• Workflow automation
• API- and webhook-connected processes
• Conversational AI
• Escalation and handoff logic
• AI onboarding and enablement
• Process improvement
• Translating business needs into practical AI systems

I’m currently building an AI-enabled SaaS product for residential real estate.

The goal is not to create another chatbot.

The goal is to create an AI teammate that can actually carry work: respond to leads, continue conversations, remember context, gather information, know when something requires human judgment, and hand the work back to the right person when needed.

A lot of my work has focused on questions like:

• What should the AI be allowed to do on its own?
• What information can it trust?
• What happens when data is missing or conflicting?
• How should it respond when it is uncertain?
• When should it ask for clarification?
• When should a response require human approval?
• When should the AI completely step out and let a person take over?
• How do you make all of that understandable to the person actually using the system?

That has led me into workflow logic, structured testing, integrations, prompts, state management, APIs, webhooks, databases, safety controls, and a lot of trial and error.

• OpenAI / ChatGPT
• Claude
• Supabase / PostgreSQL
• Twilio
• Vercel
• REST APIs
• Webhooks
• Event-driven workflows
• Next.js / React / TypeScript with AI-assisted development
• GitHub

I’m not trying to present myself as a traditional software engineer.

My strength is understanding what the business and the human actually need, then working far enough into the technology to help turn that into something real, testable, and useful.

I think AI works best as a teammate, not as a magic button.

That means giving it a clear job, boundaries, context, feedback, and a way to bring a human into the loop when judgment matters.

Good AI adoption also starts before the technology.

You have to understand the work first.

What should be automated? What should be assisted? What should stay human? What should not be happening at all?

I enjoy figuring that out.

Before spending this much time in AI, I spent years:

• helping people make high-stakes real estate decisions
• training and coaching professionals
• developing curriculum and certification preparation programs
• teaching technical and regulated subject matter
• speaking to audiences ranging from small groups to more than 1,000 people
• writing books
• hosting and appearing on radio, podcasts, video, and television
• running and supporting small businesses

That background shapes how I approach technology.

I care about whether people understand it, trust it, use it correctly, and actually get a better result because of it.

This GitHub is a public portfolio of the kind of AI and workflow work I’m doing.

The projects here will focus on practical examples such as:

• AI workflow orchestration
• human-in-the-loop escalation
• AI evaluation and testing
• API and webhook workflows
• AI adoption and enablement patterns
• sanitized case studies from larger projects

Some of my commercial product work is private, so the examples here are intentionally built to demonstrate the thinking and architecture without exposing proprietary code or data.

Thanks for stopping by.

If you're working on practical AI adoption, workflow design, customer-facing AI, or human-centered automation, those are conversations I’m always interested in.