Case Study
FLFI Fintech Platform
On a fintech platform built with NestJS, Vue 3, and EKS, I redesigned service boundaries and zero-downtime delivery so product velocity and operational reliability improved together.
- NestJS
- Vue 3
- AWS
- Kubernetes
- web3.js
Problem
FLFI was a fintech platform expanding across lending, trading, and Web3 flows. Features were growing quickly, but without clearer service boundaries and a safer release model, the operational risk would scale faster than the product itself.
When deploys are slow and rollback confidence is weak, even small changes become expensive. The core problem here was not a missing feature, but a platform shape that could not support the pace the product needed.
What I changed
I re-cut the platform around clearer domain boundaries across the NestJS backend, Vue 3 frontend, and AWS infrastructure. That made it easier to add features without each change eroding the overall structure.
On the delivery side, I built a Blue-Green pipeline with EKS, Karpenter, ArgoCD, and Helm. That gave the team a release flow with zero-downtime deploys and fast rollback, which changed deployment from a risky event into a routine operation.
At the same time, I improved key platform paths such as the Lambda@Edge + CloudFront image CDN and Web3 integration flows so reliability and latency improved together.
Outcome
The platform sustained more than 500 zero-downtime deployments over two years, and deploy frequency moved from roughly once a week to three times a day. Image delivery improved from p95 420ms to 120ms, and the team could ship with much less release anxiety.
This case mattered because it turned “platform engineering” into visible operational outcomes: better structure, safer deploys, and faster core paths.