Cyoda has set out to cut the complexity and cost of data processing and reporting with a distributed and limitlessly scalable straight-through-processing platform that delivers fast, accurate and tracked data for business use cases including regulatory reporting and internal reporting across large volumes of disparate data.
The platform challenges traditional relational database management systems and has been built from the ground up using the Cassandra open source distributed database and a technology stack developed by Cyoda engineers. Its architecture is simple by design and includes just two node types: a data node based on Cassandra, which provides an underlying, robust and scalable file system; and a processing node including a query engine, transaction orchestration, distributed reporting, messaging, workflow processing and open application programming interfaces (APIs).
Cyoda started up with support from Dassault Systèmes’ 3D FinTech Challenge accelerator programme, with which IBM is involved, and brought its platform to market late last year. The platform can be run on any number of cheap commodity blade servers and can be scaled up or down by adding or removing additional blades. It is available as an on premise or Amazon AWS cloud solution and is designed to meet the Big Data reporting requirements of Tier 1 and 2 banks.
Patrick Stanton, CEO at Cyoda, says: “The Cyoda platform is designed to take in data feeds from multiple systems and provide an aggregated view of data to other systems. Every aspect of the platform is distributed, so there is no single point of failure and no restriction on how many transactions a bank can make.”
The Cyoda model breaks away from traditional database schemas and data management approaches by offering not only scalable distributed processing, but also the ability to consume any structured source data in its existing format. Logic required to meet each new processing or reporting requirement is built only as and when needed, avoiding unnecessary data mapping and reducing time to market.
Other benefits of the platform include the provision of as-at and ad hoc reporting quickly and accurately, transactional consistency, data quality, dynamic index-based querying, data lineage that meets regulatory requirements, and a rules engine that allows non-technical users to describe the workflow of an object, perhaps a trade or legal entity. Cyoda says the platform outperforms Hadoop in many respects and that it can run simulated risk calculations roughly 9,000 times faster than traditional data processing solutions with the same hardware investment.
The company was set up four years ago by Stanton and Paul Schleger, a former colleague on a data management project at Dresdner Bank. It has since developed the Cyoda platform and is talking to a number of Tier 1 and 2 banks about potential implementation.