AI Evaluation SDET: Build Safe, Explainable AI

Jobgether, on behalf of a partner company, seeks an AI SDET based in Brazil to own and evolve an evaluation harness that tests AI agents and non-deterministic outputs. You will design automated evaluations, adversarial tests, and benchmarks while validating data quality, human-in-the-loop checkpoints, and model performance.

You will collaborate with engineering to translate findings into reliable improvements, ensuring AI systems are accurate, explainable, and compliant across diverse scenarios.