1 min lesson
Data observability & quality
Show why this lesson detail matters. "The posting asks you to ensure observability, alerting and operational standards across every layer of the data infrastructure."
Step 1 of 2
The posting asks you to ensure observability, alerting and operational standards across every layer of the data infrastructure. Written SLAs and SLOs are how you make that concrete. In the loop it becomes a probe: when a column goes null at 3am, do you find out from a dashboard or from an analyst's angry Slack?
The posting describes TB-scale first-party ingestion on a data platform that is "still early", and it asks for candidates who have scaled ingestion to billions of records a day. Treat it as a cached snapshot, because it has since left the careers board. Bad data at that scale is a production risk. In the medallion layout, bronze is raw landed events, silver is cleaned and typed, and gold is the serving layer that dashboards and agents read. One silently broken bronze table can poison a gold metric a VP reads, and the posting's charter includes data capabilities for agents, so the damage can reach agent work too. The interviewer wants a mental model of what can go wrong. Vendor names come second.