Enterprise FinTech Transformation
Modernizing a legacy banking core with Next.js micro-frontends and robust Node.js backend.
The Challenge
A leading Mumbai-based financial institution managed 12 years of transaction data on a monolithic architecture that buckled under peak load. The platform went down twice weekly during high-volume trading hours, and deploying a single feature required a full regression cycle spanning two weeks. With RBI compliance mandates tightening and customer expectations shifting to real-time banking, the legacy system was a liability.
Our Approach
ShellProp decomposed the monolith using Next.js micro-frontends and a Node.js API gateway. Each frontend module — Trading, Admin, Reports — deployed independently, enabling parallel development streams without merge conflicts. The backend adopted an event-driven pattern using Apache Kafka for transaction processing and PostgreSQL for ledger storage.
Infrastructure
Deployed on AWS ECS with auto-scaling driven by transaction volume. Redis caching cut database reads by 60%, and a custom double-write strategy migrated 12 years of data without a single minute of downtime.
Key Results
| Metric | Before | After | Improvement |
|---|---|---|---|
| Transaction speed | ~2000ms avg | ~300ms avg | 6.6x faster |
| Monthly infrastructure cost | ₹12L | ₹6.5L | 45% reduction |
| Deployment frequency | 1x/month | 10x/week | 40x more frequent |
| Critical alerts | 40+/week | 2-3/week | 93% reduction |
The platform now processes 3x the original transaction volume with headroom to scale further. The client's engineering team successfully transitioned to the new stack through a structured knowledge-transfer program.
Explore related: Backend architecture services, Cloud infrastructure projects.