What I found interesting was the fifth section that describes the strategies to improve RAG performance. Basically:<p>1. Quality control. What documents to include?<p>2. Timing. When to query?<p>3. Pre & post processing. Improve LLM outputs based on retrieved data.<p>4. End to end training. Expensive, and data intensive but possibly the best long-term approach.<p>5. Controller. An interesting idea with similarities to Reinforcement Learning.<p>I wonder what are your thoughts. Which one is most promising? What has been your experience when building RAG apps? Also, is RAG the leading architecture for building applications?