Catching silent data loss using BigQuery's native ML and Write API

2 pointsposted 11 hours ago
by boxer_shorts

1 Comments

boxer_shorts

11 hours ago

We needed a way to detect ingestion anomalies across 200+ microservices without building a separate ML stack. It turns out BigQuery already tracks every row written in WRITE_API_TIMELINE, and its built-in AI.DETECT_ANOMALIES (powered by Google’s TimesFM) can handle the detection.