Evaluation Pipelines for Experiments and Model Outputs

Typical work includes running A/B analysis, preparing experiment readouts, building eval workflows, creating repeatable scoring pipelines, and checking the quality of model outputs.