Portfolio of Ed Grzetich

How I Built a Contamination-Free API to Test AI Doc Formats

Testing how AI reads documentation requires controlling for training data. I built two custom APIs that have never existed on the public internet to isolate what actually matters.

Read More

The Problem Nobody's Measuring

Your company tracks API latency, uptime, and error rates. But nobody's tracking what it costs AI tools to read your API documentation.

Read More

Building a WebMCP Kanban Board: Browser-Native AI Integration With Zero Backend

A complete implementation of a kanban board with 8 AI-callable tools, connected to Claude through Chrome DevTools MCP. No backend. No database. No API server.

Read More

How Much Do Your Docs Cost to Read?

Did you know audience has to pay to read your docs? They do if your audience is AI. I built a tool to estimate that cost.

Read More

Claude Context Analysis: What 500 CLAUDE.md Files Reveal About How Developers Use Claude

I built a tool that scrapes CLAUDE.md files from GitHub and runs topic modeling to surface patterns in how developers configure Claude Code.

Read More

Oh My Clawd: Building a Comedy Series About AI Gone Wrong

What happens when you give an AI assistant a little too much autonomy? A comedy series exploring helpful chaos.

Read More

Making data conversational: Building MCP Servers as API bridges

How I built an MCP server as a thin bridge on top of our REST API to serve both developers and end users from a single backend.

Read More

Help panel component for AWS Amplify framework

Extending an open-source framework for user assistance.

Read More

Content review for data as a service (DaaS) design

Improving content design, usability, and accessibility for data as a service (DaaS) UX.

Read More

My Pokémon Model Context Protocol Server

Creating an MCP server for the Pokémon trading card game.

Read More