# A customer-support operations dashboard reads from Kafka customer-support events through an hourly b Canonical URL: Domain: Pipeline Design ยท Difficulty: medium ## Problem A customer-support operations dashboard reads from Kafka customer-support events through an hourly batch on the canvas. Support agents complain the dashboard is stale; their freshness target is under 15 minutes (Tier 2). Hourly batch cannot meet Tier 2; pure streaming would meet it but cost 10-50x more for what is fundamentally a 1-to-15-minute freshness need. Apply the micro-batch pattern this section just taught: replace the hourly Spark batch transform with Spark Structured Streaming or Flink (run with an explicit 1-minute or 5-minute trigger interval). Tag the warehouse mart with slaFreshness < 15min so the contract reflects the micro-batch rhythm. Do not introduce a sub-second streaming engine running in true streaming mode; the consumer accepts up to 15 minutes and a smaller trigger costs 5-10x more. ## Related - [All practice problems](https://datadriven.io/problems) - [Mock interview mode](https://datadriven.io/interview/a-customer-support-operations-dashboard-reads-from-kafka-cus-c46e5623) - [System Design Interview Questions](https://datadriven.io/data-engineering-system-design) - [Data Engineering Interview Prep Guide](https://datadriven.io/data-engineer-interview-prep) - [Daily Challenge](https://datadriven.io/daily) --- Source: DataDriven (https://datadriven.io). 100% free data engineering interview prep. Live code execution against Postgres 16, Python 3.11, and Spark sandboxes. No paywall, no premium tier, no signup gate.