Proactive Innovation Delivers $36 million in Savings
Syntel's customer, a Fortune 100 Global Financial Services Provider in Banking, Credit Cards, Travel Related and Financial Services, is one of the largest mainframe users in the industry. Their tab at IBM for mainframe CPU usage and storage adds up to over $500 million annually.
Syntel began transitioning application support and maintenance on many of this customer's systems in 2000. By 2002, Syntel teams had identified several opportunities to optimize systems to both reduce costs and improve business results. This included:
- Long running production cycles caused missing of vendor payment Service Level Agreements
- Production jobs running for hours consuming high CPU and storage
- Slow online response caused by long running database/SQL queries
The teams completed small enhancements to optimize code and storage for several of these inefficiencies. However, at an engagement level review, the Engagement Manager for the account noticed that quite a bit of work was being done in this area. Investigation was needed to determine the depth of the opportunity to save this customer millions of dollars.
After collecting baseline data from system users and examining code and performance issues across several portfolios, Syntel proposed a Demand Management solution to the customer.
In this proactive move, Syntel proposed piloting Demand Management with a two-pronged focus: optimizing code and optimizing storage.
Syntel Analysts captured initial data about the mainframe environment with a Due Diligence guide. To pinpoint opportunities for Demand Management savings, the team used "Best Programming Practices" technical guides for Cobol, JCL, DB2, SQL, IMS, VSAM, CICS with 1000+ best practices for CPU and Storage Optimization.
The following are examples of code optimization actions that were implemented by the teams.
ACTION: Using database extract instead of direct database calls
Problem => Monthly jobs accessing merchants' databases directly resulting in more CPU consumption, disputes and processing delays
Solution => Extract database into flat files and modify application programs to access data stores, eliminating database calls
Benefits to customer => $1.5 million annual savings, reduction in disputes, increased database availability, significant improvement in time to market
ACTION: Replacing database mass insert programs with concurrent BMC re-org database load utilities
POD => Daily jobs were coded to insert records into online databases.
POA => Online concurrent re-org database load utility was deployed to insert the records, removing the need for the application program
Benefits to customer => $850,000 annual savings, quick database updates delivering data view to users well ahead of processing
ACTION: Component reuse of pricing platform
POD => Similar pricing logics were being used in multiple markets using different in-house built pricing components.
POA => Syntel developed a reusable merchant pricing sub-routine capable of handling pricing for various markets
Benefits to customer = > $200,000 annual savings in development costs as well as reduced maintenance costs for multiple markets. In addition, this is saving on future development cost and improving the time to market for new pricing logics as it was built for reuse
The following are examples of storage optimization actions that were implemented by the teams.
ACTION: Decommission/merge non-performing servers
POD => Multiple servers used for similar business process
POA => Decommissioned the less utilized servers and migrated them to existing servers used for same business function.
Benefits to customer = > $2.1 million annual savings from merging server securities, reduced server maintenance costs
ACTION: Optimization of sort utility usage throughout the application
POD => Huge datasets were using IBM sort utility in conventional methods, consuming high I/O and CPU time.
POA => Optimization of sort utility usage, by specifying estimated file size parameter, resulting in less I/O and CPU time.
Benefits to customer = > $1.1 million savings annually as well as optimal utilization of existing resources
Based on the success of the pilot, the program was rolled out enterprise wide and has yielded over $36 million in savings over four years.