How I build design tools - and why process is the product...
I recently spent a day building a structured AI tool for Apple platform design. What surprised me wasn't the outcome, but the ease of decision making along the way.
Read EntryVersion 1.1
Retrieval Augmented Tumaini
Vectors from my life. They have direction, velocity, and are imbued with meaning.
The Agentic Socratic: A Skill for Intent-First Development
Most software projects fail because coding begins before the problem is fully understood. I wanted a tool that forces teams to articulate both the "why" and the "how" before writing a single line of code. Last week, I built and shipped a portable, version-controlled Socratic agent skill in a single session. It demonstrates that robust agentic workflows can be delivered as lean, executable specifications instead of complex, opaque systems.
Read Entry
The Agent-Ready Web: Building an API of Self
For decades, the web has been designed for humans—optimized for visual fidelity, interactivity, and polished user experiences. But as autonomous agents become a primary interface to information, that approach is showing its limits. When a site isn't machine-readable, it becomes invisible to AI. Content isn't merely harder to access; it's trapped behind rendering layers, JavaScript dependencies, and abstraction stacks that hide the underlying meaning from the systems trying to understand it.
Read Entry
Speed, Constraints, and Absolute Efficiency: Wiring Linear to the AI Council
For years, teams mistook process for progress, burying software development under Jira workflows, custom fields, and Agile theater. The move to Linear isn't a UI trend—it's a rejection of that complexity. Linear is built for shipping, not bureaucracy. In an AI-driven workflow, it's the only project tool that truly makes sense. Here's why—and how it powers my AI Council.
Read Entry
Ideating in the Engine Room: Why Product Strategy Requires Architectural Friction
Product management often happens detached from reality—in Miro boards, PRDs, and roadmaps where features look trivial. We sketch ideas like “User Dashboard” without ever confronting the systems behind them. That distance breeds a familiar failure mode: “How hard can it be?” The answer becomes clear only when the idea hits real architecture.
Read Entry
The Tri-Brain Architecture: Outsourcing My Memory
When building a modern digital product—juggling the stateful complexities of Next.js, schema definitions in Payload CMS, and database migrations in Supabase—the cognitive load is crushing. The natural instinct is to dump all this context into one massive LLM chat window and expect the machine to sort it out. But treating an LLM like an omniscient oracle is a rookie mistake.
Read Entry
The Auth Wall: Why AI Struggles with Stateful Systems
AI coding tools like Cursor are often sold as a straight line from prompt to production: describe the app, generate the code, deploy. In practice, building a Next.js and Payload CMS site showed something different. Development isn’t linear—it’s iterative loops with unexpected friction. Nowhere was that clearer than when setting up authentication.
Read Entry
Hello World!
For years, my professional life has centered on guiding digital products, scaling service design, and leading engineering teams to build clean, intentional software. But a corporate roadmap or a traditional CV only tells half the story. To truly understand how modern digital design is changing, you have to get your hands dirty in the machinery. This website isn't a static digital business card. It’s a living laboratory. I wanted a self-hosted, independent space to experiment with the cutting edge of modern software execution—specifically, the massive paradigm shift happening in autonomous engineering workflows.
Read Entry