evaluizer. Visualize LLM outputs against datasets, manually annotate results, and run automated evaluations to algorithmically optimize prompts.