Case study · Fintech · Payments
How Mylapay modernised DevOps to support 50M+ transactions a day.
Mylapay · Payment infrastructure
Two releases a month, deployed by hand at night, audited under PCI-DSS. We rebuilt the delivery foundation-and shipping stopped being an event.
20x
Faster deployments 2 monthly to 40+ weekly releases
25%
Lower infrastructure costs
90%
Faster issue resolution
99.95%
Uptime across regions
The opportunity
A payments platform that could clear 50M transactions, but only ship twice a month.
Mylapay runs payment infrastructure for banks and NBFCs, and the transaction rails were never the constraint. Delivery was. Releases went out twice a month, each one a manual deployment that required late-night coordination across regions, and every PCI-DSS audit cycle pulled the team off the roadmap to assemble evidence by hand.
Debugging was the same story: with no unified observability across regions, finding the cause of an issue meant working through systems one at a time. None of that is a tooling gap-it is a delivery architecture that made every change expensive, on a platform where changes are constant.
Two releases a month
Deployment was a scheduled event rather than a routine one, so work queued up behind it.
Manual, late-night deployments
Each release needed hands-on coordination across regions, timed for the quietest window.
PCI-DSS audits pulled the team off-roadmap
Frequent audit cycles meant compliance evidence was assembled by hand, repeatedly.
No unified view when something broke
Poor debugging visibility across regions turned issue resolution into a search.
The impact
Shipping became routine, and the audit trail came with it.
40+ releases a week
A fully automated CI/CD pipeline took deployment from twice monthly to 20x the throughput, with no late-night window.
25% off the infrastructure bill
Microservices on EKS and Lambda replaced provisioning that had been sized for peak and left running.
90% faster issue resolution
Grafana and Prometheus over CloudWatch logs gave one view across regions, so debugging stopped being a search.
90% less audit preparation
AWS Config rules and SSM patching made compliance evidence a byproduct of running the platform, not a project.
The stack
What it runs on.
- Infrastructure as code
- AWS CloudFormation, VPCs, EKS, RDS, S3
- Delivery pipeline
- Bitbucket, AWS CodePipeline, CodeBuild, ECR, AWS Lambda
- Observability
- Grafana, Prometheus, Amazon CloudWatch, CloudTrail, Wazuh
- Compliance automation
- AWS Config rules, AWS SSM security patching
The road ahead
The foundation is in. Next is what runs on top of it.
With provisioning, delivery, observability and compliance automated, the platform can absorb change at the rate the business wants to make it-and each further layer is an addition rather than a rebuild.
- 01
Deeper service-level observability
- 02
Continuous compliance across new regions
- 03
Progressive delivery for higher-risk changes
Talk to us
Bring one process. Leave with an outcome architecture.
30 minutes on one function-delivery, observability, or compliance evidence-mapped to the architecture behind it. No slides.