So I chose to write a CIR on StreamBase, covering all of these announcements. I usually like to get a briefing from the vendor in question before or during writing the CIR. I believe I also spoke with StreamBase during that time (though I’m not 100 percent sure---my notes are back home right now—I’m on the road covering another vendor right now). Anyway, regardless of the vendor or the announcement(s), for each and every CIR I prepare a to-myself notes doc that I call a “raw stuff.” What these notes do is boil the event down to the most concise, BS-free, description of the event (i.e., of actual substance of the vendor’s specific announcements), to serve as the starting point for my analysis. Here’s my StreamBase “raw stuff,” distilled/rewritten from the press releasese (and any other relevant info/notes in my possession) on that particular day:
StreamBase raw stuff jun 19 2007
June 19, 2007 -- StreamBase Systems announced StreamBase 5.0, the latest generation of its complex event processing (CEP) platform. StreamBase is previewing various components of StreamBase 5.0 at the Securities Industry and Financial Markets Association (SIFMA) Technology Management Conference & Exhibit in
StreamBase 5.0 includes the following enhancements:
· Eclipse-based Integrated Development Environment (IDE): StreamBase 5.0 completes the transition of the Studio IDE to Eclipse to support graphically oriented development integration between StreamSQL, Java, and Eclipse plug-ins. StreamBase 5.0 developers can extend their StreamBase CEP applications via custom operators and integrations with external systems. In addition, StreamBase application developers are now able to access and leverage the entire ecosystem of Eclipse plug-ins from within the StreamBase development environment, including plug-ins for source code version control, task management, graphical UI development and integration, XML editors, and SQL design tools.
· End-to-End Application Frameworks: StreamBase 5.0’s open development platform introduces industry-specific Application Frameworks, with the first one designed for algorithmic equities trading; this is an extension to previously announced solution sets for Reg NMS and Markets in Financial Instruments Directive (MiFID). The StreamBase Algorithmic Trading Framework is based upon a number of proven best practices for developing the full range of real-time components comprising an algorithmic trading application. The framework helps speed the development, customization, and delivery of these applications, including trading strategies, execution strategies, real-time P&L management, and real-time transaction cost analysis.
· Multi-Event CEP Pattern Matching: StreamBase 5.0 introduces enhanced pattern-matching syntax, enabling users to more easily develop applications that recognize the order, presence, or absence of complex combinations of events in real-time. Patterns can be identified within single event streams or across multiple parallel streams over any given period – whether a response is desired in real-time or over an extended time interval. With these enhancements, developers can quickly build complex event recognition applications that may be used for real-time fraud and intrusion detection, network monitoring, click stream analysis, anti-money laundering, and more.
· New Support for Advanced Data Types: StreamBase 5.0 introduces support for Binary Large Objects (BLOBs), enhancing the platform’s ability to address multimedia and document-centric application requirements, in addition to existing support for video, image, audio, XML payloads, unstructured text, and other data types.
· Expanded Data Visualization Tools for Real-Time Monitoring: StreamBase 5.0 introduces integration with several third-party visualization tools, including Adobe Flex, Microsoft Windows Presentation Foundation (WPF), Java Swing, and Eclipse Standard Widget Toolkit (SWT). StreamBase 5.0 provides bi-directional interaction with user interfaces built from these tools, enabling the creation of dashboards and other visual interfaces for monitoring and controlling real-time applications.
· High-Capacity Data Management & Persistence Framework: StreamBase 5.0 Chronicle is the second major release of the enhanced and extended persistence framework for time series data. Chronicle now provides optimized read/write integration with industry-leading, high-capacity tick stores including IBM DB2, Sybase RAP, and Vertica. In addition, Chronicle also supports bulk-loaders for these high-capacity tick stores and continues to leverage the standard JDBC interface for connecting to other historical databases.
· Enterprise-Class Security: StreamBase 5.0 includes new advanced security capabilities, including event-level security support, network data encryption, user authentication through secure integration with LDAP servers, and role-based authorization to control user access and activities.
· End-to-End Integration with External Data Sources: StreamBase 5.0 introduces the ability to use StreamBase’s Eclipse-based Adapter Toolkit to connect StreamBase CEP applications to virtually any data source. In addition, StreamBase 5.0 provides a wide range of pre-built adapters to all major market data and messaging infrastructure systems. StreamBase 5.0 adapters include TIBCO Rendezvous and
· Enhanced Administration & Run-Time Features: StreamBase 5.0 introduces improved remote administration capabilities, error management and reporting, deployment flexibility with rolling upgrades, flexible data sharing capabilities across application components, and dozens of other improvements.
In addition, StreamBase 5.0 builds on the high-performance features of previous release, and is s — and remains the fastest CEP server available today, capable of processing hundreds of thousands of messages per second per CPU. In addition, the new StreamBase 5.0 enhancements further speed the development and delivery of CEP applications that address the rapidly growing real-time processing demands of customers and partners worldwide. StreamBase 5.0 offers built-in support for IBM’s DB2 data server, WebSphere Front Office, and xSeries hardware.
StreamBase also unveiled a comprehensive CEP Reference Architecture for Algorithmic Trading for accelerating the time-to-market and increasing the extensibility of real-time algorithmic (algo) trading applications. Developed by StreamBase and Microsoft, the new CEP Reference Architecture is based on industry best practices and describes the critical real-time sections of an algo trading system and its key presentation layer components. The CEP Reference Architecture for Algorithmic Trading delivers an end-to-end framework for designing an algorithmic equities trading application and also describes the interconnection between its key real-time and user interaction components. The key real-time components are Market Data Cleansing, Data Enrichment, Trading Strategies, Risk Management, Execution Strategies, and Market Impact / Implementation Shortfall processing. Microsoft’s WPF enables rich client applications, and its powerful and flexible programming model integrates support for flexible layout, high-quality text, resolution-independent graphics, animation, video and 3D.
Furthemore, Sybase, Inc. announced that Sybase’s Real-time Analytics Platform, a highly optimized real-time data processing service platform, now integrates with StreamBase’s CEP platform. The joint solution will support real-time applications having large storage requirements, such as back-testing for algorithmic trading, risk analysis and historical trade auditing. Sybase Real-time Analytics Platform is a high performance enterprise-wide solution that delivers in-memory transaction processing, massive time-series data management, deep historical data analysis and is built on Sybase’s capital market industry-proven data management and patented data analytics technologies that have been enhanced to perform in a highly scaleable manner. Sybase Real-time Analytics Platform can deliver virtual market data feeds at accelerated speeds to match StreamBase’s high-performance CEP platform and targets a wide range of real-time front, middle and back office applications including:
- Back-testing trading strategy algorithms using virtual feeds of historical data
- Algorithmic trading applications running queries against both real-time market data and massive historical data repositories of 100+ terabytes
- Validating predictive modeling applications by comparing predicted events to actual events
- Performing pre-trade risk and compliance analysis