Why this matters
- VaultCommerce fraud-detection runs as an embedded Kafka Streams app in the payment service JVM.
- No separate cluster — scales with application pods; state in local RocksDB.
- DSL (
stream.groupByKey().aggregate()) covers 80% of VaultCommerce use cases. - Exactly-once v2 processing guarantee with
processing.guarantee=exactly_once_v2.
Source topic
groupBy key
Sink topic
Topology building blocks
KStream for record-per-event; KTable for changelog (latest per key). vaultcommerce.analytics.clicks.v1 → filter bot traffic → groupBy sessionId → count → vaultcommerce.fraud.signals.v1. Application ID vaultcommerce-fraud-detector defines consumer group and state store namespace.
Key points
- application.id — consumer group + state store prefix; unique per topology
- KStream — unbounded sequence of records
- KTable — changelog stream materialized as key-value table
- Topology — DAG of processors wired to Kafka topics
- exactly_once_v2 — transactional read-process-write guarantee
The rest of this guide is part of Pro
Free covers every fresher guide in full. Pro unlocks Senior, Staff and Architect guides across every track.
Still to come · 2 sections
- Deployment model
- VaultCommerce rollout checklist
Test yourself
Answer these before moving on — recall is what makes it stick.