Latest Articles
View all →TDD vs EDD: Why Agentic AI Needs Evals, Not Just Tests
Test-driven development assumes a stable spec and a deterministic pass or fail. Agentic AI has neither. A grounded look at Evaluation-Driven Development (EDD): what it actually proposes, what the research confirms, and the honest gap between the theory and how evals are practiced today.
The 4MAT Framework: Why, What, How, What If
Bernice McCarthy's four-question model for teaching and presenting anything without losing the room: why it beats a straight how-to, how the wheel cycles Why, What, How, and What If, and how to use it for RFCs, tech talks, and onboarding docs.
Systems Thinking for Technologists
A field manual for seeing structure instead of events: the four kinds of systems, how cause and effect actually behaves, the DART loop for diagnosis and action, and what it means for governing agentic AI.
Product Prototypes
Product Manager's Assistant
21 consulting skills on speed dial - from a raw PM query to a stakeholder-ready deliverable in seconds
How I turned 21 consulting-discipline Claude skills into a web portal any PM can use: each SKILL.md becomes a system prompt, Groq streams the answer, and the output of one skill is designed to be the input of the next.
Product Discovery Toolkit
From raw problem to validated opportunity - in your browser
A structured, PM-first discovery workflow that walks you through Ideation, Value Proposition, and Feasibility - then exports a complete evidence-backed dossier. No account needed. All data stays local.
Simple AI Agent
An AI agent in ~100 lines of Python — no LangChain, no CrewAI, just the four primitives.
How I approached a from-scratch AI agent like a product manager: framing the real problem (frameworks hide the fundamentals), running discovery, scoping a measurable MVP, and building the brain, memory, tools, and agent loop on the raw Anthropic SDK.
Smart Shop
Three AI agents. Seventeen retailers. The best price in under five seconds.
How I built an AI price comparison engine for Australian shoppers — covering the full PM journey from the fake-search problem to a parallel multi-agent architecture deployed on Vercel.
Brand Guard
AI domain intelligence that finds every brand gap before a squatter does
How I built a 7-agent AI scanner that maps a brand's global domain exposure in under 5 minutes — covering the full PM journey from problem discovery to shipped product.