Case Study
A Global Investment Manager Modernizes Fund Product Information Distribution to Increase Data Availability
The solution provides a single source of funds related to the reference, pricing, and performance data, distributed to multiple internal and external systems.
Overview
With clients in over 160 countries, this leading Global Investment Management firm provides distinct specialist investment managers, offering boutique specialization globally, bringing extensive capabilities in fixed income, equity, alternatives, and multi-asset solutions.
Challenge
The client has 80+ global websites and multiple downstream applications that consume product information. The legacy system consisted of numerous de-centralized, sometimes duplicate information stores, resulting in stale, conflicting information. It made it difficult to get to a single source of the truth, impacting public-facing website data and internal risk and compliance applications. Also, the multiple manual processes to cleanse and validate the data were expensive and error-prone.
Solution
Ness designed and implemented a data virtualization platform to source data from over 20+ sources (both internal and external), perform data cleansing, create new calculations, and make the data available through API to 50+ consuming applications. We migrated 56 product canonicals with 500+ elements each. Running on an AWS VPC, the solution involved migrating from an on-premise Oracle/MySQL database to an AWS S3 Landing Zone. From S3, the data was injected to a raw zone in Snowflake using Snow Pipes, Lambdas, and other processing units. A curation process made the data available through a special purpose zone for analytics, reporting, and consuming applications and services.
Result
The new integrated platform reduced manual processes and re-work, resulting in over $1 million in annual savings. Data availability increased from daily batch to 99% near real-time availability. Also, the platform reduced the time to introduce new data elements and reporting attributes from multiple months to a week.