Systems we built and still run
Most of our work sits behind an NDA, so what follows is the detail we are allowed to publish: the architecture, the constraint we were working against, and what changed.
Work we can talk about
Each one names the constraint we started from and the numbers we ended on.
EKS Cluster Setup for Enterprise Application
Complete Amazon EKS cluster setup with advanced networking, security, and monitoring for a large-scale enterprise application serving millions of users.
Moving an insurance ledger to AWS without stopping the business
Nine months moving core policy and transaction services off on-premises virtualisation into a multi-account AWS Organization, with Transit Gateway carrying traffic between 30+ VPCs and the remaining datacentre.
A PCI-DSS platform for a buy-now-pay-later lender
Six months building the AWS environment a BNPL lender is audited on: ECS Fargate in private subnets, Aurora PostgreSQL with read replicas, KMS envelope encryption and rotation, WAF and Shield at the edge.
One EKS control plane, many tenants, several regions
Eight months building a multi-tenant EKS platform where each jurisdiction is isolated by namespace, network policy and IRSA rather than by its own cluster, with ArgoCD ApplicationSets deploying every tenant.
500 million records a day into OpenSearch, without losing any
Four months building the ingestion and search layer for a Magento 2 estate: Kinesis Firehose buffering, Lambda cleaning records in flight, OpenSearch with rotating indices and S3 lifecycle rules behind it.
EKS for an enterprise application
Confidential, financial services
A team running containers on hand-built EC2 nodes needed Kubernetes without inheriting a platform-engineering headcount. We built the cluster in Terraform, moved the workloads across in stages, and handed back deploys their own developers run.
What people ask before hiring us
The questions that come up on nearly every first call. Anything missing, email [email protected].
Why are there so few case studies here?
Most of our work is under NDA. We publish a case study only when the client has read it and agreed to it in writing, which rules out most engagements rather than most of the work being thin.
What does a typical engagement cost?
A Well-Architected review is a fixed price agreed before it starts. Build and migration work is quoted after a scoping call, because a three-service migration and a three-hundred-service migration are not the same job. We put the number in writing before anyone commits.
How long do these projects take?
Two weeks for a review. Six to twelve weeks for a platform build like the EKS work below. Migrations run three to nine months depending on how much has to be rewritten rather than moved.
Where do you work from?
Ahmedabad, Gujarat, India, and we work with clients in their own hours rather than ours. Clients on managed operations get an on-call number that is answered outside business hours.
Can we talk to a client you have worked with?
Yes, after a scoping call. We ask the client first every time rather than keeping a reference list on standby.
Want the version with your name on it?
Send us the architecture you are running now. We will tell you which part we would change first, and whether it is worth paying anyone to do it.