IntegrationArchitect (Hardest)kafkastreamingreal-time
Architect real-time integration with Apache Kafka and Salesforce
Real World Scenario
Fraud detection requires sub-second propagation of Case updates from Salesforce to Kafka consumers scoring transactions.
Expected Answer
• Change Data Capture or Platform Events published to Kafka via middleware connector
• Kafka as enterprise event bus decoupling Salesforce from N consumers
• Schema registry Avro/JSON for event contract governance
• Partition by AccountId maintaining order per customer
• Salesforce remains producer not consumer of Kafka for this pattern typically
• Monitor end-to-end latency Salesforce change to consumer processing
• Disaster recovery: Kafka replay plus Salesforce event replay ID alignment
Follow-Up Questions & Answers
Click to expand — each follow-up includes a direct, interview-ready answer
Direct answer: Change Data Capture or Platform Events published to Kafka via middleware connector Also consider: Kafka as enterprise event bus decoupling Salesforce from N consumers In practice: Schema registry Avro/JSON for event contract governance Document the decision in an ADR and align with enterprise standards.
Architect Perspective
Kafka extends event-driven architecture beyond Salesforce limits—architects define which events exit the platform boundary.