Case Study
A leading banking company specializing in wealth and personal banking, commercial banking, and global banking and markets, with operations across 60+ countries, faced rising mainframe licensing, storage and operational costs. The bank sought to modernize its legacy landscape on Google Cloud and build a market-leading universal platform designed for 2035. The program transitioned legacy mainframe systems to a cloud-native architecture, enabling real-time applications through APIs and microservices while supporting batch workloads at scale. The target architecture combined batch processing application with a UI and real-time access layer. Curated data was exposed through microservices deployed on Google Kubernetes Engine Standard. Cloud Storage, File Watcher, BigQuery, PostgreSQL, and Apache Airflow supported data ingestion, transformation, replication, and workflow automation. The API-first design used OpenAPI 3.0 specifications with microservices developed with Java and Spring Boot. The initiative reduced development cycle time by 30% and processed 3 million records in two minutes. It also improved cost efficiency, scalability, observability, automation, and operational governance.
30% reduction
Development cycle time reduced through standardized source code, build pipelines, and deployment processes.
3 million records in 2 minutes
Processing performance exceeded SLA and NFR requirements.
Cost reduction
Legacy-system migration reduced infrastructure, licensing, and operational costs.
100% automated
Apache Airflow automated workflows and jobs, eliminating manual database operations
Improved scalability and availability
Google Cloud architecture and built-in caching improved platform responsiveness.
Key Challenges
- Rising mainframe computing, storage, licensing, and maintenance costs.
- Specialized skills increased support costs and operational dependency.
- Peak-load overprovisioning increased total cost of ownership.
- Proprietary systems constrained flexibility and technology adoption.
- Batch architecture limited support for real-time use cases.
- Legacy formats and protocols required extensive data transformations.
- Limited support for containers, orchestration, and scalable storage.
Ready to experience?
Talk To Experts
Infosys Approach
Defined services upfront using an API-first approach, OpenAPI 3.0 specifications, and an internal API design tool.
Developed Java and Spring Boot microservices using resiliency patterns, circuit breakers, bulkheads, and 12-Factor App principles.
Containerized microservices and deployed them on hardened Linux-based nodes in Google Kubernetes Engine Standard.
Used Kong API Gateway for routing, rate limiting, authentication, policy enforcement, and API governance.
Established RAW, STAGE, PROCESSED, and FINAL data layers for quality, traceability, transformation, backup, and downstream consumption.
Used Apache Airflow, Cloud Storage, File Watcher, BigQuery, and PostgreSQL to automate batch ingestion, transformation, and replication.
The Solution
Cloud-native microservices and automated batch processing on Google Cloud.
- Built a batch application for scheduled or event-driven processing of data from multiple upstream systems.
- Deployed APIs and microservices on GKE Standard as a secure, scalable, low-latency access layer for the UI.
- Standardized reusable APIs across markets and regions using OpenAPI 3.0 specifications and an internal design tool.
- Implemented hardened GKE clusters with health monitoring, node auto-repair, version maintenance, and operational automation.
- Applied data quality rules, schema checks, validation, business transformations, enrichment, and normalization across logical data layers.
- Replicated curated data from the PROCESSED layer into PostgreSQL, then promoted validated data for microservices and UI consumption.
- Defined Pod and Node resource limits and dedicated Ingress URLs to support performance, security, observability, and traffic control.
Business Outcomes
Reduced development cycle time by 30%.
Processed 3 million records in two minutes, exceeding SLA and NFR expectations.
Reduced infrastructure, licensing, and operational costs through mainframe migration.
Accelerated time-to-market with standardized DevSecOps pipelines and governance.
Enhanced observability and faster issue resolution through integrated metrics and monitoring.
Reduced manual errors by automating workflows and database operations with Apache Airflow.
Benefits
A scalable banking platform delivering faster development, high-performance processing, automation, and cost efficiency.
Reduced development cycle time by 30%.
Processed 3 million records in two minutes.
Reduced infrastructure, licensing, and operational costs.
Enabled scalable, highly available cloud-native operations.
Simplified data movement between Cloud SQL and cloud systems.
Improved visibility and accelerated issue resolution.
Reduced manual errors through workflow and job automation.
Request for services
Find out more about how we can help your organization navigate its next. Let us know your areas of interest so that we can serve you better.
