Blog

Notes from the production edge of AI systems.

Writing about quality engineering, agent reliability, and the kinds of failures teams only notice after they have already reached production.

April 20268 min readShubham Banerjee

The Invisible Risk: Why Software QA Is More Critical Than Ever in the Age of AI Agents

AI agents expand the attack surface faster than most teams expand QA. This essay explains why silent failures, leaked secrets, and unreviewed autonomous actions make quality engineering a core part of agentic infrastructure.

Read article →