Case Study
Insurance platform modernization with DDD and Kafka
Medisa ยท via Medisa
Jul 2025 - Present
Breaking a policy-and-claims platform into bounded contexts and Kafka-driven flows while baking KVKK compliance into the architecture.
- Java 21
- Spring Boot 3.x
- Kafka
- Redis
- Kubernetes
- Graylog
- Dynatrace
Challenge
An enterprise insurance platform with overloaded domains: policy, claims, and notification logic tangled together, slowing releases and blurring team ownership - all under KVKK data-protection obligations.
Approach
- Decomposed the domain into bounded contexts and independent microservices using domain-driven design.
- Designed event-driven policy issuance and claims flows on Kafka with Redis caching for real-time notifications.
- Embedded compliance into the architecture itself: response-layer masking, data classification, and audit trails designed with legal and security stakeholders.
- Built high-volume batch pipelines (hundreds of thousands of records) with transaction isolation, chunking, and index optimization.
Outcome
Improved deployability and team ownership, higher notification throughput, and KVKK compliance enforced by design rather than by convention - plus shorter release cycles via Rancher-orchestrated Kubernetes and Jenkins CI/CD.