Pitfalls in RAG Evaluation
A 0.91 faithfulness score doesn’t mean your RAG pipeline works. Most eval panels can’t see the layer that’s actually broken.
Posts
Hi, I’m Minh. I write about my experience in building and debugging agents powered by AI, as well as the errors I encounter and the knowledge I gain from them.
Picking the Right LLM Architecture
A decision framework for picking LLM architecture by asking what failure costs first and why agents are the right answer less often than you think.