"Event-Driven Trading Systems: Queues, State, and Reliable Processing in Practice"
Event-driven architectures now underpin the fastest and most resilient trading platforms in the world, yet building them correctly remains notoriously difficult. This book is written for senior developers, quantitative engineers, and architects who need to design or modernize real-time trading systems where every microsecond and every message matters. It bridges the gap between capital markets domain knowledge and contemporary distributed-systems practice, with a relentless focus on correctness under stress.
You will learn how to design a messaging and streaming backbone, model orders and risk as streams of events, and maintain authoritative state in the face of failures, replay, and reordering. Core topics include consistency models, transactions and sagas, event sourcing, stateful stream processing, and delivery semantics; practical latency engineering across networks, kernels, and runtimes; and the construction of robust OMS, EMS, risk, and market-data pipelines. The book also covers testing by record/replay, observability for trading hours, and the security, governance, and operational controls demanded by regulators.
Readers are expected to be comfortable with at least one programming language and have basic familiarity with trading concepts, though no prior experience with Kafka-style logs or stream processors is assumed. Examples emphasize implementation trade-offs, end-to-end patterns, and production realities, making this a practical blueprint for teams building event-driven trading systems from first principles or evolving legacy stacks without downtime.
