Writing
Practical essays from building React, React Native, and AI-assisted systems — architecture, workflow, and lessons from real delivery.
Featured
Vibe coding vs agentic engineering is the difference between accepting AI output on faith and keeping engineer-led control while agents do the typing. I ship production apps solo with an agentic workflow. This is how to move fast without losing the plot.
All articles
Agentic coding is when an AI agent plans, edits files, runs tests, and retries in a loop while you supervise instead of typing. Here is how the agent loop works, when to use it, and how to stay in control.
AI made building software faster. It did not make problem-solving cheaper. Implementation is a tactic; product thinking, distribution, and maintenance decide whether anyone actually uses what you ship.
You do not need to start over. If you already ship production software, AI engineering is mostly learning how to build reliable systems around models — LLM APIs, RAG, agents, and evaluation — on top of skills you already have.
AI writes pull requests faster than humans can read them. Code review cannot stay line-by-line inspection. The scarce resource is judgment: what belongs in the product, what gets automated, and where senior attention actually matters.