Case Study : Data Integration
problem
The USA Patriot Act of 2001 required financial institutions to implement anti-money laundering procedures that analyze all transactions for suspicious, fraudulent or abusive activity. This banking client was facing stiff financial penalties for non-compliance.
Process
Meridian Technologies was engaged to resolve the clients AML crisis and was given a six-month window to complete and deploy a solution. Meridian immediately applied their data integration, analyst and testing services. Utilizing the Ab Initio platform, Meridian was able to quickly integrate the clients 26 systems of record. The data produced by these systems included over 30 million daily transactions. During this process Meridian applied sophisticated data modeling and pattern detection algorithms which worked in concert with the AML business rules to identify new types of suspicious, fraudulent and abusive transactions.
From there, the data was integrated into load-ready formats: (i.e., customer information, account information, cross reference and transactions), which integrated with the bank’s existing transaction systems. To automate the operations for the solution, over 200 jobs with cross dependencies were developed. Comprehensive test scenarios were also created to account for all possible operational situations and verify the system’s ability to detect new types of anomalous transactions which were not specifically included in the business rules.
These scenarios were validated through extensive Systems Integration Tests (SIT) and User Acceptance Tests (UAT). Finally, all system enhancement and additions were thoroughly documented and knowledge was transferred to internal MIS staff. The completed solution was deployed under budget and significantly ahead of schedule. The AML data integration system framework continues to be the “gold standard” for system and application development and deployment at this client.
Results
Successful completion of the AML project allowed this banking client to avoid severe financial penalties. The deployed data analysis application was engineered to support many applications beyond AML where benchmarking and base line analysis are required to identify trends in large data sets. This allowed the client to recognize benefits beyond the original scope of work. Although the client’s plan included a five-member team to provide ongoing AML support, the deployed system required the support of only one offshore resource. Since its deployment, the system has experienced zero production failures.
Case Study : Finance Data Warehouse
problem
This top-five bank commissioned a large initiative that would realize cost savings by consolidating its finance data warehouse and tax-data repository onto one infrastructure. The solution needed to accommodate 130 new data sources, new undefined data elements and consolidation of two reporting systems into common data structures.
Process
Meridian designed the architecture, data model and business processes and delivered a reference implementation of the architecture for use by subsequent development efforts. We began by creating a conceptual architecture that logically depicts the acquisition of data, new common data, legacy application processing and data provisioning. We then created the functional architecture and design, technical specifications and delivered the solution.
Results
Meridian delivered the architecture, design and code, which included end-to-end workflows, logical and physical data models (with over 7,000 data elements), ETL data acquisition and integration, and general ledger reconciliation processes to certify the data for use, YTD roll-up calculations for fees.
Case Study : EDW Strategy
problem
This large financial-services company had been on a multi-year journey to replace several line of business data marts with an integrated enterprise data warehouse (EDW) that could serve the reporting-and-analysis needs of the finance, credit-risk and marketing areas of the bank. The client had made a significant investment in time and money working with other solution providers in failed attempts to deliver an integrated data platform for the enterprise.
Process
The client engaged Meridian to assess their current state, capture their highest priority business requirements, and recommend a viable, cost-effective strategy to deliver accessible, standardized, and authoritative information to the enterprise. Meridian worked collaboratively with the client’s finance, credit-risk, marketing and technology experts, facilitated discovery workshops, assessed the efficacy of current platforms, and proposed the future-state solution.
Results
Meridian proposed the approach, target architecture and roadmap to deliver the strategic data management platform for this client. The roadmap included a phasing strategy, defined business and technical priorities, and cost estimates to deliver the future state solution. The project culminated in presentation of the detailed business case, and alternative strategies, to the client’s executive team. The client ratified the recommended strategy and funded the first phase of projects.
Case Study : Performance Engineering
problem
This banking client developed a platform called eFMA, a .NET based web front-end to a commercial master data-management system. The development team was experiencing performance problems while processing bulk data uploads. Each record uploaded required over 1.5 seconds for processing, requiring over four hours to process a 10,000 record file.
Process
The client engaged Meridian Technologies to perform a brief (two weeks) performance engineering engagement.
Results
Meridian provided easy-to-implement recommendations that reduced the record-processing time to about 0.5 seconds per record on average, which tripled the processing speed and reduced the runtime for 10,000 records to about 1.5 hours.
