Databricks' KARL agent uses reinforcement learning to generalize across six enterprise search behaviors — the problem that breaks most RAG pipelines.

Databricks has released KARL, an RL-trained RAG agent that it says handles all six enterprise search categories at 33% lower cost than frontier models.

Are AGENTS.md files actually helping your AI coding agents, or are they making them stupider? We dive into new research from ETH Zurich, real-world experiments, and security risks to find the truth ...

Agent skills shift AI agents toward procedural tasks with skill.md steps; progressive disclosure reduces context window bloat in real use.