Field notes on building AI agents, RAG systems and LLM applications that hold up in production.https://leverge.aien-ushello@leverge.aihello@leverge.aihttps://leverge.ai/insights/building-an-llm-evaluation-sethttps://leverge.ai/insights/building-an-llm-evaluation-setThe practical method — where the cases come from, how to adjudicate them, and how to validate a model-as-judge before you trust its numbers.Thu, 16 Jul 2026 00:00:00 GMTevaluationevaluationLLMOpstestingqualityhttps://leverge.ai/insights/why-ai-agents-fail-in-productionhttps://leverge.ai/insights/why-ai-agents-fail-in-productionSix failure modes that account for almost every agent project we have been called in to rescue, and what actually prevents each one.Thu, 02 Jul 2026 00:00:00 GMTai-agentsAI agentsproductionevaluationreliability