MetaFluent for Value-Added Data

Trading organizations today develop an increasing number of applications that ingest dynamic data such as streaming prices and execution reports, and publish data streams containing rates, analytics, positions, and risk information. While it would be nice to use legacy technology for this integration, that can be a challenge. Such solutions involve many moving parts with complicated configuration, leaving the solution fragile and with poor performance.

The MetaFluent Managed Content Suite provides a simpler and faster way to create value-added data streams. MetaFluent integration and transformation is in-process rather than involving multiple components distributed across a network. Underlying data from multiple sources are integrated via inline adapters. Data models are pre-configured via metadata, provide consistency and control and can include any sort of transformation or enrichment. Examples are integrating real-time and reference data or augmenting content using in-line analytics, such as portfolio valuations.

Analytics can be pre-calculated or instantiated dynamically in response to a user request. For example, MetaFluent can supply particular ranges of a yield curve to each user without needing to calculate unobserved ranges. MetaFluent can also manage collections such as portfolios, order books, and data chains on behalf of applications, enabling developers to refer to these collections as single entities. MetaFluent can also integrate with CEP (complex event processing) systems such as Esper for more complicated algorithms.

Behind the scenes, MetaFluent provides functions critical for dynamic data, such as integrating with leading entitlements systems, managing initial values and updates, notifying clients of potential stale data, and synchronizing derived and underlying data.

Co-locating this functionality in a single process yields both simplicity and higher performance. Yet this does not limit scalability. MetaFluent supports multiple clustering paradigms, including the ability to implement topic-based clustering. In this mode, an administrator can align each analytic with the appropriate resources (e.g., hardware, data), minimizing resource usage and maximizing responsiveness.