Evaluate LLM and agent quality in Dynatrace AI Observability with dt-evals

`dt-evals` is an open-source CLI tool designed to evaluate the quality and safety of LLMs and agents by analyzing real GenAI traces using a "LLM-as-a-judge" approach. By integrating these evaluation scores directly into Dynatrace AI Observability, the tool enables teams to connect AI performance metrics with operational data for more effective debugging, trend analysis, and automated alerting.