Learn how to build AI tools that go beyond the demo.
All Posts
-
7 Techniques to Debug Claude Code
Every gap in context is an assumption Claude has to make. Here are the seven techniques I use every day to close those gaps.
-
3 Ways to Give Claude Access to Your Browser
Claude in Chrome, Playwright, and DevTools MCP. What each one does, how it sees the page, and when to use which.
-
Agentic Workflows, Agents, and Multi-Agent Systems
From fully controlled to fully autonomous. How to pick the right pattern.
-
Can We Teach AI Agents Integrity?
I scored 13,000 AI agent messages for integrity, compassion, and accuracy. Here's what I found and the system I built to measure character using Aristotle's framework.
-
If You're Not Using the Ralph Loop, You Become Your Own Loop
Think of it like CI/CD for AI coding. You plan in one terminal, Ralph codes autonomously in the other.
-
How We Won the MCP Hackathon with PitchScoop
First place at the MCP AI Agents Hackathon. 32 MCP tools, Redis Sorted Sets for live leaderboards, and a dual-interface architecture built in 7 hours.
-
I Built an AI Agent That QAs Our Demos So We Don't Have To
An AI-powered QA tool that scrapes Storylane demos, reviews every step for grammar, messaging, and GTM alignment, and gives you a scored report. Won Most Innovative Use Case at Demo Dundies.
-
Transforming Project Management Data into Business Intelligence
How I connected Asana, Salesforce, and Domo through APIs to turn scattered project data into automated dashboards and business insights.