Proving the concept for faster month-end processing and building a platform for Big Data.
Analysing terabytes of data for this IBM client’s month-end financial close could take up to five days – limiting the ability of the finance team to make timely strategic decisions at the start of each new month.
The company worked with IBM and Northdoor on a successful proof-of concept to accelerate month-end processing with the IBM® PureData™ System for Analytics, powered by Netezza® technology.
Results suggest the month-end process can be cut by 50 hours, enabling financial close within a day. Should offer lower TCO than alternative vendors’ solutions and deliver full ROI in three to five years.
About our Insurance client
A leading market insurer needs to close its books at the end of each month – a complex process involving the analysis of four terabytes of data from the company’s operational data store and underwriting systems.
Although the company had a robust data warehousing environment and a well established set of ETL (extract, transform and load) and analytical processes, the sheer volume of data made the monthly close a very lengthy process – often taking as long as five days.
“Until the monthly close is completed, it’s difficult for the business to understand its current position,” explains a spokesperson for the insurer. “This was putting a limit on our activity at the start of the new month. If we could accelerate the process, our finance team would be able to make timely decisions and release funds to the business more quickly, making us more agile in the market.”
The company began to consider its options for accelerating its month end processes. A traditional approach to improving performance – such as adding more database servers, replacing the existing IT landscape with more powerful hardware, or re-designing the process itself – would be expensive and disruptive. Instead, the company decided to investigate the possibility of migrating its data warehouse onto an analytics appliance.
The company considered offerings from several leading vendors, and realised that only IBM PureData for Analytics, powered by Netezza technology, could meet its requirements. Most of the alternatives were either too large and expensive, or lacked key capabilities such as schema migration or in-database analytics. By contrast, the PureData solution was projected to deliver a full return on investment within three to five years.
“It was important to find a solution that could work with the schema of our existing data warehouse, as redesigning it would have made the migration too complex and costly,” comments the spokesperson. “We didn’t want to have to retrain our IT staff, so easy manageability was critical too.
“We also wanted a solution that would integrate well with our existing ETL tools, as well as the IBM Cognos® Business Intelligence software that will soon become our strategic analytics platform. Over the longer term, we plan to use the appliance for more than just month-end processing, so we wanted the most versatile solution available.”
Embarking on a proof-of-concept
The company decided to run a proof-of-concept exercise to demonstrate the value of the PureData solution and help to build the business case for investing in it. Working with IBM and Northdoor, an IBM Platinum Business Partner that specialises in solutions for the insurance sector, the in-house team migrated more than 3,500 database tables onto the PureData appliance, loaded four terabytes of data into them, ported over 100 database procedures (comprising 60,000 lines of code), and repointed 400 ETL mappings.
During this process, PureData compression technologies were able to reduce the total size of the data from four terabytes to just one, and query performance on the new appliance was improved by a factor of between 20 and 150, depending on the type of query.
Once the PureData appliance is up and running in the company’s production environment, it should be able to host development, test and integration environments as well as production. This will enable the company to retire or re-purpose a number of its existing servers and storage systems – cutting the complexity of the data warehousing landscape, and reducing data centre and management costs.
The IBM PureData System for Analytics, powered by Netezza technology, gives this leading market insurer a simple, easy-to-deploy, pre-optimised analytics platform that will help to dramatically accelerate the processing of huge volumes of information – enabling the company to reduce its month-end close process, and laying a foundation for future Big Data analytics initiatives.
The success of the proof-of-concept, and the strong business case that the project team was able to build to support the adoption of the PureData solution, persuaded the company to go ahead with a full-scale deployment, which is now in progress.
“If we see the same results in production as we did in the proof-ofconcept, we expect the PureData appliance to pay for itself within three to five years,” concludes the spokesperson. “Moreover, the month-end close process is just the first of many areas where we see PureData’s potential. As we move forward, we’re interested in using it for other types of big data analytics too, such as statistical analysis, data-mining and geospatial analytics, which will really help us get the maximum value from our investment.”