Beyond Accuracy: Engineering Trustworthy AI Systems in Production

As AI systems become increasingly integrated into high-stakes, real-world applications, the challenge of ensuring their reliability, transparency, and accountability has come to the forefront. Traditional model evaluation metrics such as accuracy and F1 score are insufficient to capture the complexities of system behavior in production environments. This panel, “Beyond Accuracy: Engineering Trustworthy AI Systems in Production,” brings together senior practitioners across domains including machine learning, data science, observability, and compliance. Drawing on their experience with large-scale deployments in financial services, enterprise software, and data infrastructure, panelists will examine methodologies for validating AI systems beyond model performance, discuss strategies for monitoring data and model drift, and explore frameworks that integrate governance, explainability, and auditability into the AI lifecycle.