Open to senior AI solutions architect & engineer roles
I design enterprise AI that holds up in production: RAG and agentic systems with the guardrails, evaluation, and human-in-the-loop controls that make them safe to ship.
Cincinnati, Ohio area
See them live
Not slideware. Each is a working, AWS-native product you can click through, built to show how I handle grounding, agents, trust, and the jump from prototype to production.
Live on AWS
Turns high-risk cases into cited, evidence-backed recommendations, with a human on every decision.
The case studies
Every project leads with the business problem and is honest about the trade-offs, the governance, and the cost.
An AI platform that turns fragmented, high-risk support cases into cited, evidence-backed resolution recommendations, and keeps a human in control of every decision.
Problem
Enterprise support teams sit on scattered case data and tribal knowledge, and most AI demos ignore the governance reality of acting on a real customer account.
A governed multi-agent system that drafts evidence-backed answers to security questionnaires, catches unsupported commitments before they become liabilities, and routes anything risky to a human.
Problem
Security questionnaires stall enterprise deals for days, and an AI that auto-answers them will confidently commit a company to things it cannot back up.
An interactive AWS Solutions Architect (SAA-C03) playbook with 118 topics, decision guides, and a grounded AI tutor that answers only from the material it can cite.
Problem
Most cert prep is passive recall. The harder skill, and the one the exam actually tests, is reasoning through architecture trade-offs under real constraints.
How I work
The same approach whether it's a proof of concept or a production system — the goal is always something a team can confidently own.
Before any technology, get specific about the outcome that matters and how it's measured.
Map who is affected, who decides, and the measurable results that define success.
Choose an architecture that fits the problem — not the other way around — and name the trade-offs.
Prove the risky parts early with prototypes that have clear, agreed success criteria.
Security, governance, cost, monitoring, and adoption — the difference between a demo and a system.
Expertise
Grouped by where it matters: the models, the data foundation, the cloud they run on, and the work of helping people adopt them.
RAG systems that stay grounded in sources, and multi-agent workflows that know when to hand a decision back to a person.
The unglamorous parts that decide whether an AI system ships: guardrails, source attribution, human approval, and an audit trail.
Serverless, event-driven AWS designs built for cost, scale, and least-privilege security from the first diagram.
The customer-facing craft: discovery, solution design, and explaining hard trade-offs to engineers and executives alike.
A little about me
My work sits at the point where an AI idea has to become something a team can actually run. I like designing RAG and agentic systems, and I spend most of my attention on the parts that decide whether they ship: grounding, guardrails, evaluation, and keeping a human on the decisions that count.
I have built production ML systems, led a data science team, and spent the last stretch customer-facing. Lately I have been building my own AWS applications, which means I care as much about whether a solution gets adopted, and stays safe, as whether it works in a demo.
Whether you're hiring, exploring a proof of concept, or trying to get an AI initiative to production — I'm glad to help think it through.