Project Showcase

RapidLaunch: Orchestration That Actually Configures

Case Study

RapidLaunch:
Orchestration That Actually Configures

A multi-cluster, provider-agnostic deployment platform that doesn’t stop at launching servers — it configures them, secures them, and manages their entire lifecycle.

Applied Minds AI
·
Infrastructure Orchestration
The short version

Most orchestration tools launch containers and hand you an IP address. Between “container exists” and “service is production-ready” lies a gap filled with package installation, network configuration, firewall rules, DNS records, secret injection, and security hardening. RapidLaunch closes that gap entirely. Multi-cluster, provider-agnostic, with billing integration (WHMCS now, Blesta and HostBill coming) and zero password exposure via HashiCorp Vault.

The Problem: “Deployed” Is Not “Ready”

Most orchestration tools spin up a container or VM, hand you an IP address, and call the job done. But a launched server isn’t a configured server. Between “container exists” and “service is production-ready” lies a gap filled with:

  • Package installation and system configuration
  • Network setup — interfaces, bridges, VLAN assignment
  • Firewall rules — UFW, iptables, default-deny policies
  • DNS records — A, PTR, MX, all pointed at the right IPs
  • Secret injection — API keys, database credentials, DKIM keys
  • Security hardening — Fail2ban, Auditd, AIDE, SSH lockdown
  • Service setup — start applications, verify health checks
  • Monitoring — Prometheus, Grafana, alerting rules
  • Backups — encrypted, scheduled, verified

That gap is where most of the operational time goes — and most orchestration platforms ignore it entirely.

What RapidLaunch Does Differently

RapidLaunch is a multi-cluster, provider-agnostic deployment platform that doesn’t stop at provisioning. It actively configures every server it launches — from package installation and network setup to DNS record creation and secret injection. An automated DevOps system with AI support, managed from a single control plane.

31K
Container capacity — atomic VMID allocation, race-condition safe
30s
Provision to production-ready — packages, networking, hardening
0
Exposed credentials — Vault manages every secret in the stack
<50ms
API response — async job submission, 5 concurrent workers

Multi-Cluster, Multi-Provider

One API, unlimited clusters. OVH, Hetzner, Vultr, colocation — any Proxmox-compatible infrastructure, managed from a single control plane. The intelligent placement engine handles capacity-based selection, geographic affinity, and compliance constraints (HIPAA, GDPR cluster isolation). Migrate infrastructure between providers without changing a single API call. No vendor lock-in.

Full Lifecycle Management

Provision, configure, suspend, unsuspend, terminate. Containers go from “ordered” to production-ready in 30–60 seconds. The API supports lookup by container ID, VMID, or external reference (e.g., whmcs:12345) — so billing systems can track resources using their own identifiers.

Zero Password Exposure

Every credential in the stack is managed through HashiCorp Vault. Encrypted at rest. Runtime retrieval with automatic token renewal. Full audit trail. AppRole authentication with least-privilege policies. No secrets in environment variables, config files, or process lists. The system was designed from the start with the assumption that every other layer will be compromised — Vault is the last line of defence, and it holds.

Billing Integration

The API is billing-agnostic by design. The orchestrator never sees pricing — billing portals translate customer tiers into resource requirements, and the orchestrator provisions accordingly. WHMCS addon shipping now with full admin and client area integration. Blesta and HostBill modules in development. Any billing portal can integrate via the REST API.

The Stack

Flask REST API behind Nginx (HTTPS-only, localhost binding — Flask never exposed to the network). Celery workers for async processing with Redis (AOF + RDB persistence, no job loss). PostgreSQL for persistent state and atomic VMID allocation. HashiCorp Vault for every secret. SSH to Proxmox hosts for container operations. Ansible playbooks for configuration management.

What “actively configures” means — an example

When RapidLaunch provisions a CompleteSender instance, it doesn’t just create a container. It installs KumoMTA, configures the Lua routing layer, sets up the Go backend, deploys the React UI, configures Nginx with SSL, installs the full security stack (UFW, Fail2ban, Auditd, AIDE, RKHunter), deploys Vault Agent for secret management, configures Prometheus + Grafana monitoring, sets up Restic backups, and runs health checks. All automated. All in under two minutes.

<img>
Diagram: RapidLaunch Architecture
Three-tier diagram. Top: Billing portals (WHMCS, Blesta, HostBill) connecting via REST API. Middle: RapidLaunch Orchestrator (Flask API, Celery, PostgreSQL, Redis, Vault). Bottom: Proxmox cluster farm with multiple clusters and LXC containers for CompleteSender (70000-79999), PowerDNS (80000-89999), Postal (90000-99999). Side: Gitea, SonarQube, Ollama cluster.
Suggested: 1200×800 · Architecture diagram · Dark background, colour-coded tiers

Need infrastructure orchestration that goes beyond “launch”?

Multi-cluster, provider-agnostic, fully configured. From container creation to production-ready in under a minute. With billing integration and zero password exposure.

Book a Strategy Call →

Interested in what we build?