Axisforge
CASE STUDY 03 • PLATFORM ENGINEERING

From manual deployments to a Kubernetes platform serving 200 engineers

The Situation

An e-commerce firm was experiencing frequent deployment delays due to inconsistencies across staging and production. Their 200-person development team depended on complex scripts and manual console steps to publish applications, slowing down update releases.

What We Found

  • Environments drifted frequently as a result of manual CLI modifications.
  • Permissions management lacked granular IAM policies, creating security exposure risks.
  • Deploy pipelines lacked test verifications, causing errors to reach production.

What We Did

We built a GitOps platform using Amazon EKS, ArgoCD, and Terraform:

1. Infrastructure as Code

We wrote Terraform configuration blocks to manage the cluster provisioning, network routing, and IAM credentials.

2. GitOps Delivery

We set up ArgoCD to coordinate application updates, using version-controlled git pull requests as the source of truth.

3. Developer Templates

We created standardized application configurations, enabling engineers to set up new environments independently.

The Outcomes

Deployment success rates rose significantly, and average environment creation time was cut from 3 days to under 10 minutes.

Engagement Metrics

Client Profile
Large E-commerce Platform
Core Services
Platform Engineering, Kubernetes
Timeline
8 Weeks
10m
Deploy Time
200+
Devs Enabled

Engineering Reflection

Building the platform required balancing standard controls with developer flexibility. If we did this project again, we would spend more time configuring initial local build templates to reduce validation steps during testing.

Planning a Kubernetes transition?

We work with B2B SaaS and technical firms that demand high engineering standards. Let's discuss your cloud setup.

Start a conversation →