Cloud & Hosting Platforms

Our client needed to compete in the crowded cloud hosting market with a differentiated platform that could onboard indie developers and small teams without the steep learning curve of AWS or GCP. Existing white-label solutions were either too rigid to customize or too expensive to license, and rolling a completely custom platform from scratch seemed like a multi-year endeavor no lean startup could afford.

The core technical challenge was multi-tenancy at the infrastructure level. Each user workspace needed full isolation — separate containers, distinct networking, scoped secrets — while keeping resource utilization efficient enough for the business model to work at ₹500/month entry pricing. The platform also had to handle SSL provisioning for custom domains automatically, without human intervention, and support zero-downtime deployments so that users never experienced an outage during deploys.

On the product side, the client wanted a dashboard that could rival Vercel and Railway in polish: real-time build logs, usage analytics, a built-in code editor for quick config changes, and team collaboration features including role-based access control. The scope was ambitious for a team of three, and the timeline was tight — four months to a public beta.

We chose a microservices architecture built on Kubernetes, with each tenant's workload running in its own namespace. For the orchestration layer, we wrote a custom controller in Go that watched for deployment events and managed the full lifecycle: pulling code from Git, building Docker images via Buildpacks, pushing to a private registry, and rolling out new versions with health checks. This controller became the backbone of the platform's CI/CD system, replacing what would otherwise require stitching together half a dozen third-party services.

For SSL, we integrated cert-manager with Let's Encrypt and built a DNS verification proxy that could handle wildcard certificates for the platform's shared domain as well as single-domain certs for custom domains. The entire provisioning flow — from a user typing their domain name to HTTPS being live — took under ninety seconds on average. We instrumented every step with structured logging so failures could be diagnosed in seconds rather than hours.

The frontend was a Next.js application with a real-time WebSocket layer for streaming build logs and deployment status. We used a combination of server-side rendering for the marketing pages (critical for SEO) and client-side rendering for the dashboard (critical for interactivity). State management was handled through Zustand with optimistic updates, giving the UI a snappy feel even on slower connections. The analytics module pulled metrics from Prometheus and displayed them through custom chart components built on top of Recharts.