Writing
Notes on building software, tools and learning AI engineering.
-
Lessons from building a coding agent package
What building Brief Context taught me about safe project scanning, adapter design, and giving AI coding agents better handoff context.
-
How I added free article narration with Piper, then moved to Kokoro for quality
A practical look at adding local, free audio narration to a static Astro site, what worked with Piper, and why Kokoro became the better quality option.
-
Resist the temptation to prompt
A reminder for AI-assisted development: before asking the model for code, slow down long enough to understand the problem, boundary, and next smallest useful step.
-
How I evaluate AI-assisted code before merging
A practical review checklist for AI-assisted development: intent, tests, edge cases, architecture, security, and maintainability.
-
My AI-assisted development workflow without cutting corners
How I use AI coding tools for research, scaffolding, tests, and refactors while keeping architecture and review human-owned.
-
Cloudflare Workers architecture patterns I keep reusing
Patterns for building practical serverless products with Workers, Hono, D1, R2, Queues, and Durable Objects.
-
Building Gorgi: a one-line grounded AI chat widget for websites
How I designed a multi-tenant AI chat widget with tenant isolation, quota checks, and a tiny Shadow DOM embed.
-
How I design human-in-the-loop AI systems
A practical checklist for AI workflows that need approvals, audit trails, spend limits, and safe resume behavior.
-
Inside Flare Control: orchestrating AI agents on Cloudflare's edge
A practical architecture tour of a queue-driven, human-in-the-loop AI agent platform built on Cloudflare Workers.
-
From idea to shipped mobile app with Expo, maps, and serverless
Lessons from building JigSpot: offline maps, GPS tracking, React Native, Expo, MapLibre, and a Cloudflare backend.
-
The Cloudflare AI stack I would use again
A practical architecture breakdown of Workers, Hono, D1, R2, Queues, AI Gateway, and React for AI-enabled products.
-
AI features that should not be chatbots
Why many AI product ideas are better as inline suggestions, drafts, classifiers, or workflow automations instead of chat interfaces.
-
Remote work with AI tools: faster is not automatically better
How distributed teams can use AI coding tools without losing shared context, review quality, or product judgment.
-
What product engineers should learn in the AI era
The skills that matter for product-minded engineers as AI changes how teams prototype, build, and evaluate software.
-
An AI learning roadmap for working software engineers
A practical AI learning path for engineers who need to build useful product features, not become researchers overnight.
-
How I scope a product build before writing code
The questions I use to turn a product idea into a buildable plan, from user loop to architecture, risks, and first milestone.
-
Why full-stack engineers matter more on AI product teams
AI product teams need engineers who can connect models, interfaces, data, workflow, cost, and user trust into one working system.
-
What maintaining production frontends teaches you
Lessons from production frontend work that matter in real products: performance, accessibility, state boundaries, and long-term maintainability.
-
Documentation is remote team infrastructure
Why durable docs, decision records, and clear setup notes matter as much as tools for distributed software teams.
-
Turning ambiguous requirements into shipped software
How senior engineers reduce ambiguity, define milestones, and turn vague product ideas into working software.
-
Remote engineering is mostly written communication
How strong async writing helps distributed teams make better technical decisions with fewer meetings.
-
Async code review habits that make remote teams faster
Practical code review habits for distributed engineering teams, from smaller pull requests to better reviewer context.
-
Hello, world
Why I built this site and what I plan to write about.
No articles match this filter.