Next Generation Trade Surveillance System Using In-Memory Computing Technologies
There has been a growing trend to adopt In-Memory computing technologies to address the following scenarios:
- Real-time insights into business
- Supporting business growth specially in banking & capital market
- Digital paradigm such as IoT, machine learning, streaming analytics
This is an abstract of work done in designing a trade surveillance system for a large stock exchange demanding a throughput of 1M messages per sec with low latency. The requirements included processing of streaming data, alert calculations & In-Memory data store for online reporting. The talk will take your through a detailed evaluation of various solution options to address above requirement. Findings of Apache Ignite will be discussed in detail which is the core component of our solution. This includes design considerations, optimization, and performance benchmarks.