Large Refactors with AI, with Humans Still in Charge
Refactoring with AI on large codebases: make it safe first, codemods vs AI, small verifiable batches, reviewing AI diffs, and when to keep AI out.
Writing
Refactoring with AI on large codebases: make it safe first, codemods vs AI, small verifiable batches, reviewing AI diffs, and when to keep AI out.
Treat prompts like code. Use DeepEval to write test cases, score faithfulness, relevancy, and hallucination, and run LLM regression tests in CI before you ship.
React Native developer roles and responsibilities by level: what a senior engineer owns in architecture, releases, quality, native code, APIs and people.
A fair React Native take-home assignment: principles, a copyable brief, a scoring rubric, the follow-up interview, a senior variant and an AI tools policy.
A React Native folder structure that scales: thin Expo Router routes, feature folders, shared primitives, one API layer, config and when to use a monorepo.