Strategist · Engineer · Operator
Engineering intent
in the era of AI‑augmented software.
Three decades building software, leading teams, and shipping product — from AOL and Microsoft's first .NET era to founding Alekto and leading the blockchain practice at Smashing Boxes. Now back at Microsoft, writing about how engineering fundamentals become more essential, not less, as agents enter the loop.
Now Writing
The Intent‑Engineering Field Guide
Series · Article 01 · Start here
Gherkin May Be the Most Consequential Language of the AI‑Augmented Era
Coding agents ship code faster than any human ever has — but without careful guidance, they make codebases worse. The discipline that wins isn't typing speed; it's the ability to specify intent precisely enough that an agent can act on it.
What I Work On
Three lanes, one thesis: fundamentals matter more, not less.
01 / Intent Engineering
Specifying behavior agents can act on
Behavior-driven development, Gherkin, contract-first thinking, and the disciplines that make AI-generated code worth keeping. The current writing focus.
02 / Strategy & Architecture
Three decades of building product
Microsoft Enterprise Strategy, founding and running Alekto, leading the blockchain practice at Smashing Boxes, MIT Sloan executive education in platform strategy and design thinking.
03 / Blockchain & Sovereignty
Be your own bank
Long-running interest in financial sovereignty, non-custodial wallets, and the cultural side of crypto. Documented on the Block Boys YouTube channel.
The Codified Practitioner
Your expertise is the raw material. The methodology is the process for getting it out.
Senior engineers have a problem most AI training doesn't address: their expertise lives in their heads where AI can't reach it. The Codified Practitioner is an eight-stage program for converting accumulated engineering judgment into AI systems that can execute against it reliably.
Learn about the program →Surface
Make tacit expertise visible
Specify
Express it as AI-executable intent
Build
Implement working AI systems
Demonstrate
Produce a portfolio that proves it
Code isn't cheap. Bad code is the most expensive it has ever been. The teams who win this era will be the ones who can specify intent well enough that their agents can be trusted with the keyboard.
Selected Writing
A curated archive.
One in Twenty: A Founder's Story
How a personal experience with credit reporting errors led to founding Alekto — and what the lean-startup grind actually looks like from the inside.
The Importance of Abstraction in Cloud Computing
On why platform abstraction is the durable bet — written in 2009, when "cloud" still meant something specific. The argument has aged better than expected.
Architectural Fidelity with Globally Distributed Teams
Risk management for software architects working with offshore teams. The principles — contracts, fidelity, governance — apply directly to AI agents now.
Code Contracts Help Distributed Teams
A fifteen-year-old argument for design-by-contract that turns out to be exactly the right primitive for working with coding agents.
So…, Bail-Ins Are a Thing
On a quiet form of consumer banking risk and why non-custodial wallets exist. One of the foundational pieces behind the Block Boys channel.
Block Boys — Be Your Own Bank
Financial-literacy education and blockchain explainers. Why depositing money makes you an unsecured creditor, and what to do about it.