Problem
- Daily and monthly processing of customer data and contracts on the suite is performed based on a myriad (>400!) of different batches
- In addition, two PEAK days per month require increased runtime due to the need to process all contracts
- At the earliest, the end-of-day processing may take place after the POS is closed and must be completed before midnight, i.e., a maximum of four hours is available
- By discontinuing various legacy systems and transferring their data to the central system, the number of contracts there will double within a few months and subsequently even increase tenfold
Contribution ResultONE
- Setup of a task force with the involvement of all suppliers and the operating teams
- Organization of the appropriate test environment and test data
- Initiation of the target-oriented analysis, optimization and database tuning measures
- Introduction of twice-a-day meetings for coordination, the provision of deliverables, and clarification of the next steps
- Daily provision of information to various stakeholders such as program management and shareholders
Result
- Despite doubling the number of contracts, significant improvements in various key performance indicators were achieved, including
- Halving the daily batch processing time
- Processing the volumes of PEAK days in 2.5 hours
- As part of a preparatory deployment, extensive mechanisms for comprehensive load balancing and the flexibly parameterizable, highly scalable parallelization of batch processes were implemented
- Other future improvement opportunities were developed and documented