Skills

I design high‑load services end‑to‑end: boundaries, data ownership, contracts, and the operational model. I’m comfortable with strict performance requirements and make pragmatic trade‑offs to keep systems predictable and maintainable.

I build and evolve microservice platforms and integrate internal banking systems with external providers. I focus on stable contracts, fast partner onboarding, and safe rollout of new flows without breaking existing consumers.

I design REST/SOAP/GraphQL APIs that are easy to consume and hard to misuse: versioning, idempotency, and clear error models. I care about documentation, backwards compatibility, and “sharp edges” that become production incidents later.

I build reliable async pipelines with Kafka/RabbitMQ: retries, dead-letter queues, ordering constraints, and consumer safety. I treat “at-least-once” as a reality: idempotency, deduplication, and monitoring of lag/throughput are part of the design.

I work with PostgreSQL/Oracle/ClickHouse and approach performance systematically: indexes, query plans, batching, and hot-path refactoring. I use Redis caching when it’s justified and measurable—reducing DB load without creating consistency surprises.

I build real‑time services that stay stable under load: thousands of concurrent sessions, backpressure, and resilient reconnect flows. The goal is consistent latency and predictable behavior—not just “it works on my machine.”

I build services that are observable by default: meaningful metrics, structured logs, traces, dashboards and actionable alerts. I’m comfortable with incident response and postmortems, turning failures into concrete engineering improvements.

I implement authN/authZ with OAuth2/OIDC, Spring Security and Keycloak: least privilege, secure defaults, and threat-aware design. I pay attention to sensitive data handling (PII), auditing, and safe failure modes.

I ship reliably: automated pipelines (GitLab/Jenkins), containerized workloads (Docker), and orchestration (Kubernetes/OpenShift). I care about safe rollouts, quick rollback paths, and reproducible environments across dev/stage/prod.

I write maintainable Java/Kotlin: clear naming, simple abstractions, and refactoring that keeps velocity high over time. I rely on unit/integration tests for critical flows and use reviews as a tool to prevent tech debt from accumulating.