arize-prompt-optimization - Optimize LLM prompts using Arize trace data
Extracts prompts from Arize traces, analyzes evaluations and annotations, and generates improved prompts through a data-driven optimization loop.
Tags
Updated: 2026-10-05Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Extract prompts from spans
- Inspect chat messages
- Analyze evaluation feedback
- Analyze human annotations
- Compare outputs with ground truth
- Optimize prompts with LLMs
- Export traces and datasets
- Inspect experiment results
Inputs
- Arize project
- Arize space
- Trace data
- Span identifiers
- Dataset name
- Experiment name
- Evaluation results
- Human annotations
- LLM provider credentials
Outputs
- Extracted prompts
- Prompt analysis
- Revised prompts
- Exported trace files
- Exported dataset files
- Exported experiment files
Requirements
- ax CLI
- Configured Arize profile
- LLM provider credentials
- Supported LLM model