Technical Product Leader | AI Development | Enterprise Architecture
25 years building developer tools. Recently modernized a 610,000-line codebase using spec-drive AI-assisted development, while enhancing the product with built-in agentic tools. From Sun Microsystems to serving major clients like the US Bureau of Economic Analysis and the New York Marathon, I bridge the gap between customer problems and shipped products.
Writing about AI-assisted development, legacy modernization, and practical software engineering.
Building a natural-language reporting system for personal finance using AI. Reverse-engineered the Copilot Money API and created a Chrome plugin with MCP server capabilities and an agentic workflow.
At Web Performance, I've served enterprise clients including the US Census, Canadian government, and New York Road Runners.
Personally implemented a zero-downtime modernization of a 25-year-old, 610,000-line Java codebase using the Strangler Fig pattern and AI-assisted development. Integrated AI with Amazon Bedrock and AgentCore, Migrated from Java 1.4 to 11, Eclipse RCP 3.6 to 4.19, and replaced the legacy SWT desktop UI with a modern React front-end.
Built entirely from scratch using Claude Code to address a key customer need: side-by-side comparison of load test results across different test runs. Demonstrates systematic AI-assisted development from spec to deployment.
End-to-end project built with Claude Code, Github SpecKit and multi-LLM-model collaboration. Uses a comprehensive set of CI/CD tools with 800 unit tests, ~135 integration tests, pre-flight validation, auto-snapshop before deployment, pre-commit validation, etc.
Natural-language financial reporting system using Claude Code and MCP server architecture. Reverse-engineered Copilot Money's GraphQL API and built a Chrome extension with agentic workflow for complex analytical queries.
A niche labor of love: a CircuitPython library for driving scrolling LED displays showing real-time data from anywhere. Powers the physical LED product sold at ThemeParkWaits.com.
Technical deep-dive and development approach
I've got a bunch of software ideas that have been on the back burner, unimplemented, because they would have taken weeks full-time to implement, and who has that much spare time? When I came across GitHub Speckit a few months ago, I wanted to test it on a complicated personal project before risking it at work, and I've been wanting to do this one for ages. A twist is that I integrated the Zen MCP server into the dev process, so it uses multi-LLM consensus for design and implementation details.
"Theme Park Hall Of Shame" tracks ride downtime and wait times across major US theme parks and calculates "shame scores" for unreliable attractions. There are plenty of sites that track wait times; the point here is to provide industry-wide insights and analysis. The interesting part isn't the app itself—it's the development guardrails I've built into the Claude Code workflow.
The front-end was iteratively designed with Claude Desktop. The original prompt was to create a website theme based on the work of Disney legend Mary Blair, best known for creating It's a Small World and widely considered the most influential of Disney's conceptual artists.
The next thing to add is pattern recognition, once there's more than a couple of weeks of data. I'm interested in seeing whether there's any relationship between weather and reliability.
Main Dashboard
Performance Charts
Wait Times Heat Map
Shame Scores by Park
Performance Metrics