https://www.montecarlodata.com/blog-ai-agent-evaluation/
AI Agent Evaluation: 5 Lessons Learned The Hard Way
Alik Peltinovich, Michael Segner, Elor Arieli
Oct 15 2025
"The unique evaluation challenge for AI agents... is that they are non-deterministic.... This means a traditional CI/CD evaluation framework leveraging tests with explicitly defined outputs doesn’t work for agentic systems."