Independent notes on engineering and organizations

Software, AI & People

Long-form essays on software engineering, artificial intelligence, organizations, and people

A developer carries an open laptop everywhere to watch approvals, while another leaves the workstation in the studio and handles the task calmly from a phone
Latest essay Why I Built PandaForce

All essays

9

Why I Built PandaForce

Software Engineering · Why I built PandaForce as a self-hosted web workspace for supervising Codex, working with files where they already live, and eventually fitting AI into enterprise systems.

Can Software Craft Be Passed On in the Age of AI?

Software Engineering · From The Art of Computer Programming and Refactoring to Design Patterns and The Mythical Man-Month, an essay about rebuilding the feedback loops through which software craft is learned.

Will AI Coding Kill Low-Code?

Software Engineering · AI coding weakens the old low-code speed advantage, but makes proven components, runtime contracts, and enterprise governance more valuable. A practical argument for combining natural language, Spark CLI, and a trusted application platform.

Revisiting Peopleware: The Better AI Gets, the More Software Depends on People

Software Engineering · A fresh reading of Peopleware for the age of coding agents: when code is abundant, attention, trust, judgment, and shared responsibility become the scarce resources.

Rereading The Mythical Man-Month: Software Engineering in the Age of AI

Software Engineering · A rereading of The Mythical Man-Month for the age of coding agents, context engineering, and knowledge-driven software development.

Reliability by Design: What AI Still Misses in Production Systems

Software Engineering · AI can generate complete features quickly, but it rarely designs the failure path by default. A practical framework for critical flows, fault boundaries, idempotency, observability, and recovery.

AI Coding Makes DFX More Important, Not Less

Software Engineering · AI can implement features remarkably well, but it does not automatically produce a system that is reliable, maintainable, observable, secure, and safe to operate. DFX turns those qualities into explicit design constraints.

Accelerating Business Innovation: Lattice and Clean Architecture in Practice

Software Engineering · How Lattice contains business change at scale, and how its architecture principles became an executable Skill for AI-assisted development.

How Should a Programmer Measure a Year?

Organizations · An essay first written in April 2018 about measuring an engineer’s growth beyond promotion—through craft, influence, judgment, responsibility, and sustained learning.

Topics

→
Software Engineering8Organizations1

About

I’m Rocky Yu. I write about software engineering, AI, and how people and organizations build systems that work in the real world.