Real-Time Innovations (RTI) and Collaborative Software Initiative (CSI) have partnered to offer a standards-based offering for market-data distribution. The companies paired RTI Data Distribution Service middleware with low-latency CSI Feed Handler.

RTI’s middleware avoids servers, daemons and extra network hops. CSI handles the feeds like OPRA, ARCA and ITCH. The combination removes critical latency from the front end of the trading system, the companies claim.

According to the companies, RTI and CSI leverage the standards-based offering for distributing market data, the Open Market Data (OMD) architecture. Reportedly, OMD was developed collaboratively by financial institutions, technology providers, service providers and middleware vendors.

In addition, RTI and CSI have entered into a collaboration to develop a Feed Handler plug-in for OMD that uses the standard Data Distribution Service (DDS) application programming interface (API) to enable integration and deployment of market-data platforms. This plug-in offers a standardised configuration to adapt systems to demanding requirements such as scaling, customised reliability, performance tuning and application fusion.

Another open API in the OMD architecture, Market Data Abstraction Layer (MDAL) provides a standardised way for an application to access market data, while the use of DDS enables distribution of market data to SQL databases, Complex Event Processing, data recording applications, and Microsoft Excel spreadsheet software, the companies added.

Sumeet Shendrikar, director of technology, Financial Services for RTI, said: Operational risk and deployment costs have become significant factors when selecting technology. With budgets becoming tighter, time-to-deployment cannot be sacrificed by spending cycles on the complex integration of proprietary interfaces. CSI Feed Handlers paired with RTI Data Distribution Service reconcile both needs: relentless improvements in performance and latency, and the ability to integrate and maintain the system in the long run.