How to Build Reliable AI Products: A Production Engineering Guide
AI demos are easy. Reliable AI products are an engineering problem.
Technical posts, product decisions, and honest observations — written while building, not after.
The best place to start if you want to understand how I think and work.
AI demos are easy. Reliable AI products are an engineering problem.
MV3 is four moving parts — manifest, service worker, content scripts, and UI — not a single script tag anymore.
Stop presenting projects. Start shipping a system that compounds trust.
Build in public
What shipped, what broke, and what changed — written as it happened.
Notes now have a clear cross-post path for LinkedIn and X, plus copyable post text. The goal is simple: publish once on the site, then drive attention back to the owned page.
ResistGate and Amethyst now have explicit demo sections using real screenshots, with room to replace them later with GIFs or recorded walkthroughs.
Wrote three long-form pieces targeting the gaps I see in production AI work: reliable product design, agent architecture patterns, and prompt engineering for real inputs. These are the things I keep explaining to people. Now they are linkable.
What AI features look like when regular people use them in real life.
Use the interface that matches who's reading — humans, software, or agents.
Start with the data model, then expose the same truth through pages, APIs, MCP tools, and examples.
Practical guides for Manifest V3, privacy-first browser tools, and extension distribution.
The Web Store review is a product gate — privacy policy, assets, and MV3 compliance matter as much as your code.
Real products need friction by design — ResistGate makes you type before distraction, not just block URLs.
Pricing, product choices, and what I've learned trying to make things people actually buy.
A personal site should not only be browsed by people. It should also be legible to the tools people use to discover, evaluate, and remember work.
Your website should not only be visited. It should be callable.
Architecture tradeoffs, primitives over complexity, and durable system patterns.
A technical launch is easier to distribute when the story, snippets, and proof points are already packaged.
The first rule of agent-readable systems: do not give the agent more power than it needs.
How I make decisions, stay focused, and keep moving without a team.
PayPal subscriptions, webhooks, and a Postgres entitlements table — the smallest stack that unlocks Pro in the extension.
Freemius, Paddle, and custom PayPal each fit a different MRR band — pick the stage you're in, not the stack you wish you had.
Sleep, routines, learning, and what actually keeps me at full capacity.
Filter notes by topic and month.
Topic
Month