Data Orchestration

The reality is that, in most enterprises, multiple systems house mission-critical data – data that can’t be easily shared with other systems throughout the enterprise. Historically, most strategies for centralizing data management for multiple, disparate systems fall into one of three categories:

• Data warehouses - Using ETL (extract, transform, load) tools to consolidate normalize, and load data into a centralized location for analysis.
• Golden copy systems - Contributing to and querying from a centralized master set of reference data.
• Federated data systems – Enforcing rules under which each system notifies other systems of changes in data.

Each approach has serious limitations. Data warehouses excel at historical analysis, but are severely limited in providing real-time access. The golden copy approach is fraught with organizational dependencies and control issues and doesn’t scale well. And federated data systems can quickly become unmanageable due to an overwhelming number point-to-point connections and dependencies.

Aleri offers an innovative solution to data orchestration that overcomes the hurdles of traditional approaches. Aleri uses an event-driven architecture that leverages Complex Even Processing (CEP) for rules-based transformation and intelligent routing, for continuous synchronization of data across heterogeneous systems.

The Real-Time Enterprise

Aleri CEP forms part of an Event Driven Architecture that extends traditional Service Oriented Architecture (SOA) with the paradigm of events. As individual systems make changes to data sets, they send asynchronous notification events to a data management hub running on an Aleri server. The Aleri server uses CEP technology to apply transformation and processing rules to update other dependent systems.

The Aleri CEP architecture is uniquely suited to the challenges of data orchestration because of its ability to maintain state – in other words, it can apply intelligence to determine which systems need to be updated based on its knowledge of the state of each system. It can also apply rules that control what data can be shared as well as when updates should take place.

This intelligent data hub architecture for data orchestration offers several advantages over the traditional methods:

• Real-time capabilities - Multiple systems can stay synchronized as changes occur, rather than relying on batch updates.
• An event-driven architecture - Each system only broadcasts changes and receives notifications — versus communicating with every other system on a one-to-one basis.
• Centralized integration and transformation rules – The protocols for sharing information are managed in one place, rather than in dozens of point-to-point connections.

For an in-depth look at how CEP-based data orchestration is being applied in the securities industry, download this free report from the Aite Group, titled, “EDM to DDM: Moore’s Law Rewards Centralized Data Procrastinators.”