US Automobile Giant Improves Mainframe Batch SLA from 32% to 99%

Fortune 100 US Automobile Manufacturer

• The client's mission-critical mainframe batch ecosystem was missing SLAs more than two-thirds of the time, throttling downstream manufacturing systems and delaying month-end financial close.
• Years of incremental JCL changes had created an opaque, fragile dependency graph where any single failure required overnight manual recovery — driving up MIPS costs and operational risk.

• Kumaran ran a full discovery sweep across thousands of JCL members, datasets, and scheduler triggers to rebuild an authoritative dependency map and prioritize the highest-impact jobs.
• We migrated batch logic from JCL/COBOL to Spring Batch on Java, moving execution off the mainframe to a distributed Linux runtime — preserving exact business semantics through a parallel-run validation harness.
• Performance engineering and intelligent re-sequencing of the job graph removed serial bottlenecks, while observability dashboards gave operations real-time SLA visibility for the first time.