I build applied AI systems — tools, workflows, and agents that actually ship.
Applied AI engineer who goes beyond prototypes. I design, build, and deploy AI-powered systems — from real-time voice agents to retrieval pipelines to workflow automation. My background in learning systems means I build for real users, not just demos.
Systems I've built and shipped
Not prototypes. Not slide decks. Working AI systems deployed in production — with real users and measurable outcomes.
AI Voice Agent Simulator
Real-time voice simulation using OpenAI Realtime API. Agents practice customer conversations against a lifelike AI opponent — deployed, used daily in production training.
RAG Knowledge Navigator
Retrieval-augmented generation pipeline that surfaces the right information at the right time. Built for real-time agent support during live customer interactions.
AI Workflow Automation
Automated end-to-end workflows using AI coding agents — content generation, quality review, deployment pipelines. Cut production time by 60%.
AI Coach & Feedback System
Automated performance coaching that analyzes interactions and surfaces actionable improvement suggestions. Closed the feedback loop without human review bottlenecks.
Watch me build it, not just talk about it
Every tutorial is a real build of a real system. No theory, no slides — just working code and the decisions behind it.
Building a Real-Time Voice Agent with OpenAI Realtime API
Walk through how I built a production voice simulation tool — from API setup to streaming audio to deployment. Not a tutorial app — a real system.
Automating Workflows with AI Coding Agents
How I use Codex and other AI agents to automate end-to-end workflows — from generation to review to deployment. 60% time reduction, no hype.
Building a RAG Pipeline That Actually Works in Production
Step-by-step: chunking strategies, embedding selection, retrieval tuning, and the stuff nobody talks about — like latency budgets and failure modes.
Applied AI, built for real people
I build AI systems that ship — voice agents, retrieval pipelines, workflow automation. Not prototypes. Not slide decks. Working systems with real users and measurable outcomes.
My edge: I spent 13+ years designing systems for how people actually learn, adopt tools, and change behavior. Most AI engineers build for specs. I build for humans. That means the tools I ship actually get used — because they're designed around real user behavior from day one.
I think about AI the way I think about any tool: does it solve a real problem, and will someone actually use it? If the answer isn't clear, I build a small experiment first. If it works, I ship it. If it doesn't, I learn and move on.
What I Build With
Let's build something
If you need someone who can design, build, and ship applied AI systems — not consult, not pitch, but actually build — let's talk.