
Broker readiness review
We audit topic partitioning, retention, and consumer lag alarms for Kafka, RabbitMQ, or cloud-native buses you operate in Taiwan or hybrid regions.

Project-based programming consulting for teams shipping event-driven software—bounded contexts, async contracts, and broker topologies that stay operable after launch.
Discuss a scope brief: +886-2-0000-1203
Before we write a line of integration code, we verify your checklist: domain events are named by business facts, consumers declare schemas, and failure paths are replayable without silent drops.
Engagements are fixed-scope programming projects—not open-ended staff augmentation. Each quote lists deliverables tied to brokers, handlers, or outbox tables you already run.

We audit topic partitioning, retention, and consumer lag alarms for Kafka, RabbitMQ, or cloud-native buses you operate in Taiwan or hybrid regions.
Time-boxed implementation of transactional outbox or inbox patterns against your existing RDBMS, with chaos tests on duplicate delivery.
Schema registry rules, compatibility gates, and CI checks so event payloads cannot break downstream services without an explicit version bump.

Usually no. We integrate with the broker and observability stack you already pay for, then document topology changes only when throughput or ordering requirements demand it.
Yes. Workshops run in English with async follow-ups; pairing sessions target handlers, serializers, and deployment manifests in your repository.
Runnable code, replay runbooks, contract tests, and a written decision log explaining trade-offs on ordering, sagas, and dead-letter handling.

Event backbones we design must remain observable under production traffic—not only in diagram form.