Netflix’s catalog metadata plays a vital role in the streaming experience, and any errors can have immediate consequences for millions of users. A recent incident highlighted a significant shortcoming in Netflix's approach to data validation. While their code canary system successfully detected issues in code deployments, it failed to catch a corrupted data feed that rendered metadata for certain titles empty. This resulted in playback failures, exposing a gap in their resiliency strategy for data pipelines, which had not been held to the same rigorous standards as code. To address this gap, Netflix developed the Data Canary Orchestrator Pattern, a system designed to validate catalog metadata using production traffic while minimizing customer impact. The solution involves a dedicated orchestrator instance that coordinates validation flows, maintaining two service clusters: one for the current production catalog and another for new versions under review. This setup allows for the running of chaos experiments to test new data in real-time. The orchestrator checks the health and synchronization of both clusters before triggering these experiments, ensuring any issues are detected and contained quickly. Meeting a strict 10-minute validation timeline required enhancements to their chaos platform. Netflix customized experiment thresholds, allowing them to detect failures more effectively based on Starts Per Second (SPS), a direct measure of customer impact. They also implemented sticky canaries that keep user traffic consistent during experiments, avoiding cross-contamination of results. The system streams metrics in real-time, aborting experiments immediately if regression is detected, which prioritizes speed over statistical confidence. This approach reinforces their commitment to maintaining a high-quality viewing experience.