Worked on internal core banking web application powering hundreds of branch tellers and multiple retail banking workflows, including SLA-critical end-of-day regulatory reconciliation where reliability and auditability were mandatory.
Migrated 50+ legacy Java 6 services supporting teller operations and regulatory reporting to Java 11, eliminating legacy failure paths while preserving RBAC, auditability, and regulatory SLAs during phased production rollouts.
Defined service boundaries while decomposing 20+ tightly coupled SOAP services backing core customer onboarding workflows into REST APIs, eliminating cross-team deploy dependencies and enabling independent rollouts and safe rollbacks.
Owned end-to-end design of 30+ SLA-critical Spring Batch reconciliation pipelines processing 20-30 GB/day across millions of financial transactions/day, using idempotency and checkpointing to prevent duplicate settlements.
Replaced freeze-window deployments with containerized CI/CD pipelines (Docker, Kubernetes), enabling incremental rollouts, fast rollbacks, and 40% shorter release cycles for SLA-critical services.
Prevented partial writes and settlement job failures during peak windows by correcting PostgreSQL schema drift uncovered in production.
Primary on-call engineer for core banking services during production migrations, resolving Sev-1/Sev-2 incidents, executing rollbacks, and restoring service availability within SLA windows (1-3 hours).
Software Engineering Trainee
LTIMindtree
Hyderabad
08.2021 - 10.2021
Contributed to migrating a production banking datastore from Oracle to PostgreSQL, focusing on schema redesign and index optimization to improve query performance by 40% for internal infrastructure services.
Optimized 100+ latency-sensitive SQL queries used by internal banking services through execution-plan analysis and indexing, reducing response times by 20% under peak load.
Worked on transactional data reconciliation workflows during staging cutovers, validating cross-schema consistency and preventing data integrity issues prior to production rollout.