Amit Rao

For 3 years, as part of Project NeMo (Network Modernization) I've been part of an extremely complex transformation of American Express card processing network from Mainframe Monolith to Java-based Distributed microservices architecture

Lead Agile Scrum team and oversaw integration, scaling, optimization, and secure deployment of logical microservices using Kubernetes orchestration handling upwards of 100 Million transactions everyday

· Team Lead: Directed a high-performing scrum team, spearheading strategic planning and agile execution with business and product stakeholders.

· Global Migration:Entrusted with managing four services and an end-to-end data synchronization job during American Express's critical migration from monolith to Java microservice architecture. Oversaw a 36-hour migration where any failure would have reversed the entire global migration.

· REST API SOR: Designed a scalable REST APIproject using Vert.x with RxJava for asynchronous, non-blockingoperations, and Guice for lightweight dependency injection. Achieved active-active multi-DC setup with Couchbase XDCR and GTMfor traffic management. Secured with nginxload balancing, IDaaS, and mTLS between issuers.

· Data Sync Job:Led cross-team development of a Sparkproject synchronizing 50 million daily transactions between India and global data centers. Scheduled via Ansible and YARN, processing Hadoop Parquetsevery 30 minutes, impacting 20 services. Improved inter-team collaboration, mitigated bugs, and ensured timely delivery.

· Envoy Proxy Resolution:Resolved a complex, year-long production bug in Envoy(L7 proxy) affecting retry mechanisms under full load in multi-pod scenarios. Fixed circuit breakertripping due to simultaneous in-flight retries when downstream services failed unexpectedly. Upgraded Envoy API from v2 to v3 during resolution.

· Bespoke Settlement Report:Developed custom settlement reports for network partners using Spark, Hadoop Parquet, and Elasticsearch. Optimized Spark job by removing `row_number` function, implementing eager caching, and reducing custom sorting logic, reducing processing time from 3 hours to 57 minutes.

· Cold Storage Service for Reports: Led end-to-end development of a Spring Boot microservicefor long-term report storage in Amazon S3 (up to 7 years), allowing fetching PDF/Excel reports based on user authorization. Used Cassandra(active-passive setup) for report metadata. Handled deployment, CI/CD pipeline, and Consul/Vault updates.