Enterprise AI coding fails on people, docs and process — not the codebase

_jaydeepkarale · x · 2026-11-16

jaydeepkarale argues the codebase is the least problematic part of enterprise AI coding — the real blockers are people lacking knowledge, poor documentation, and horrible processes.

The quoted post notes you can't just ask AI to write code and push to production in large enterprise codebases: established coding standards, shared files, dependencies, predefined workflows and architectural patterns all require deep understanding, which is why software developers aren't going anywhere anytime soon.

Original post →

More from coding & agent

coding & agent channel →