UK private client investment manager Brewin Dolphin has implemented Zoomix’s Accelerator enterprise data management platform to automate the matching, standardization and classification of hundreds of thousands of customer and financial data records.
Accelerator is a so-called ‘self-learning’ data management platform that analyzes the relationships and meanings relevant to any data domain by observing decisions made by business users and incorporating them into its knowledge base for application to future operations.
The idea is to avoid the lengthy and expensive development of rules, scripts and dictionaries required by other approaches to data management, while adding accuracy and flexibility in complex data environments.
For Brewin Dolphin, the implementation is straddling management of both financial instrument and customer data. Brewin Dolphin is integrating Zoomix Accelerator into its data processing infrastructure for the online processing of both data types.
In addition to providing the software, Zoomix is providing consulting services to aid in the integration of the platform with Brewin Dolphin’s existing IT infrastructure and to train business users.
For financial instrument data, Accelerator will be used to consolidate large volumes of data from external sources such as Thomson and Reuters (now Thomson Reuters) into the company’s operational databases while ensuring that all records remain consistent, complete and unique.
For client data, Accelerator will provide the platform for Brewin Dolphin to build a solution that will help it assign each client a unique customer identification, regardless of how many accounts or recorded name variations existed for the client. This included joint accounts, with the added complexity of multiple names assigned to individual accounts. This requirement grew out of U.K. Financial Services Authority (FSA) transaction reporting obligations.
Brewin Dolphin manages portfolios worth £21.6 billion for over 120,000 clients. Brewin Dolphin Investment Banking is adviser to around 131 corporate and 100 institutional clients.
Brewin Dolphin selected Zoomix Accelerator for its self-learning capabilities and its ability to process both customer and financial product data on a single platform. It was also the only solution analyzed by Brewin Dolphin that offered all three of the data processing functions sought: matching, standardization and classification.
According to Aurora Diaz, divisional director of data management at Brewin Dolphin, “We were impressed with the software’s ability to deliver a high degree of accuracy in a very short period of time, without any development effort on our part.”
Jerusalem-based Zoomix develops and markets software to accelerate the implementation of master data management solutions. The Brewin Dolphin contract appears to be its first major win in the financial services segment; previous announced contracts have focused on more generic data management issues in such sectors as health care, energy and entertainment.
Zoomix Accelerator is a scalable automated data processing server that solves data inconsistencies in-line with business processes. Rather than attempting to try to cleanse and correct data before it can be considered fit for business needs, Zoomix Accelerator resolves conflicts among data entities appearing in different systems in real time. This alleviates the need for what Zoomix describes as “expensive and marginally successful ‘off-line’ data correction projects.”
Its proprietary self-learning technology combines “advanced semantic analysis with machine learning to automatically classify, match and standardize any kind of corporate master data.”
This, according to Zoomix, allows it to deal with all types of semi-structured, complex corporate data, including highly variable product and financial data. Its patent-pending Self-Learning Technology allows the software to learn, understand and correct any type of complex master data on the fly, it says.
This Self-Learning Technology allows users to create an initial usable master data set within days, rather than months or years. By observing user behaviour, the technology ‘learns’ how to automatically aggregate, combine, standardize and classify data from multiple sources. Henceforth, data correction is addressed automatically, with continuously improving accuracy.
This automatic interpretation of data means that “Zoomix Accelerator never requires the development of scripts or the manual definition or maintenance of data processing rules,” the company says. The system automatically interprets data to learn the optimal matching, normalization and classification required for any data domain.
Zoomix says Accelerator works equally well with financial data, customer data, supplier and product data delivered in any language or combination of languages. The system contains embedded technologies – such as automatic attribute extraction – to process both structured and semi-structured data.
Finally, the system’s integration with daily user workflow processes provides on-the-fly data correction. Erroneous, duplicated and incorrectly categorized data are repaired immediately, Zoomix says, without disrupting routine business functions.